279 69 13MB
English Pages 1196 Year 2012
English Historical Linguistics HSK 34.1
Handbücher zur Sprach- und Kommunikationswissenschaft Handbooks of Linguistics and Communication Science Manuels de linguistique et des sciences de communication Mitbegründet von Gerold Ungeheuer (†) Mitherausgegeben 1985−2001 von Hugo Steger
Herausgegeben von / Edited by / Edités par Herbert Ernst Wiegand Band 34.1
De Gruyter Mouton
English Historical Linguistics An International Handbook
Edited by Alexander Bergs Laurel J. Brinton Volume 1
De Gruyter Mouton
ISBN 978-3-11-020220-5 e-ISBN 978-3-11-025159-3 ISSN 1861-5090 Library of Congress Cataloging-in-Publication Data A CIP catalog record for this book has been applied for at the Library of Congress Bibliographic information published by the Deutsche Nationalbibliothek The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available in the Internet at http://dnb.dnb.de © 2012 Walter de Gruyter GmbH & Co. KG, Berlin/Boston Cover design: Martin Zech, Bremen Typesetting: Apex CoVantage Printing: Hubert & Co. GmbH & Co. KG, Go¨ttingen ⬁ Printed on acid-free paper s Printed in Germany www.degruyter.com
Preface to Historical Linguistics of English
The study of the English language has a lengthy history. The second half of the 18th century saw a phenomenal increase in the number of published grammars of the vernacular language, while the field of comparative linguistics arising in the 19th century was concerned in large part with the Germanic languages, including English. However, it is in the field of theoretical linguistics that English has played a truly central role. While there are no reliable statistics, it seems safe to say that the majority of studies in contemporary linguistics deal at least in part with English, and are also written in English. During the 20th century, monumental works concerned with the English language, both synchronic and diachronic, were produced, following historical/comparative and more contemporary linguistic approaches. In keeping with developments in the field of general linguistics, today it is possible to find descriptions and analyses of the history and development of English from virtually any linguistic perspective: external, internal, generative, functional, sociolinguistic, pragmatic, comparative, phonological, morphological, syntactic, lexical, semantic. There are numerous “Histories of English” to cater to just about every (theoretical) taste, as well as detailed descriptions of historical periods, language levels, or theoretical frameworks of English and specialized studies of individual topics in the development of the language. Work on the history of English has culminated most recently in the seven-volume Cambridge History of the English Language, edited by Richard M. Hogg (1992–2001). Study of the history of any language begins with its texts. Increasingly, however, scholars are turning to dictionaries and corpora of English that are available online or electronically. The pioneer historical corpus of English, the Helsinki Corpus of English Texts, was first released to scholars in 1991. The third edition of the Oxford English Dictionary online is now fully integrated with the Historical Thesaurus. The searchable Middle English Dictionary, completed in 2003, is available online along with the Corpus of Middle English Prose and Verse. The Dictionary of Old English Web Corpus is also searchable online. ARCHER, A Representative Corpus of Historical English Registers 1650–1990, accessible at a number of universities, provides a balanced selection of historical texts in electronic form. COHA, a 400-million-word, balanced Corpus of Historical American English 1810–2009, was launched online in 2010. Smaller corpora, such as the Corpus of English Dialogues 1560–1760, the Lampeter Corpus of Early Modern English Tracts, the Corpus of Early English Correspondence, the Corpus of Early English Medical Writing, and the Old Bailey Corpus, have made more specialized corpora available to scholars. Archives of historical newspapers online, including the Zurich English Newspaper Corpus, provide another source of electronic data. Finally, syntactically annotated corpora for historical stages of English are being produced, including the The York-Helsinki Parsed Corpus of Old English Poetry, The York-Toronto-Helsinki Parsed Corpus of Old English Prose, The Penn-Helsinki Parsed Corpus of Middle English, and The Penn-Helsinki Parsed Corpus of Early Modern English.
vi
Preface to Historical Linguistics of English Taking into account the important developments in the study of English effected by the availability of electronic corpora, this Handbook of English Historical Linguistics offers a comprehensive, interdisciplinary, and theory-neutral synopsis of the field. It is meant to facilitate research by offering overviews of all the relevant aspects of the historical linguistics of English and by referring scholars and students to more indepth coverage. The handbook is intended primarily for researchers in the field of (historical) linguistics generally, as well as for researchers in allied fields (such as history, literature, and culture). The handbook comprises two volumes, each volume consisting of approximately 70 articles written by a wide variety of authors from a number of different countries world-wide, representing a variety of theoretical approaches, and including both younger scholars as well as more established experts.
Volumes 1 and 2 The sequencing of material in the two volumes of the Handbook of English Historical Linguistics is bottom-up, beginning with detailed studies of the periods, levels, and linguistic components of each period. The second volume moves to a higher level, with a focus on general underlying concepts, theories, and methods as well as new and hitherto rather neglected approaches to the history of English. While the two volumes form a set, with cross-reference as far as possible in order to facilitate reader-guidance, they are also capable of standing alone. Following this essentially inductive approach, then, the first volume (edited by Laurel J. Brinton) is focused on the details of English language history. After overviews of the recognized periods of English (Section I), the volume then treats the linguistic levels. These are broadly understood to include newer components such as prosody, pragmatics, phraseology, discourse, styles, registers, and text types as well as more traditional areas such as orthography and onomastics in addition to the fully acknowledged areas of phonology, morphology, syntax, and semantics (Section II). These summaries will be useful both to students and to those not working directly in the field of English historical linguistics, such as typologists. Sections III–VI contain detailed descriptions of the different periods – Old English, Middle English, Early Modern English, and Late Modern English – in respect to the range of linguistic levels; discussions of language contact, standardization, sociolinguistics, and literary language are included for most periods. Moreover, for each period, selected important phenomena (such as the development of do-periphrasis, the Great Vowel Shift, pronoun usage, or relativization) have been chosen for more detailed study. Following the treatment of the different periods, the volume addresses a variety of questions of standardization (Section VII), such as the effects that dictionaries, the Bible, language attitudes, and codifiers have on normalizing the language. The last section (VIII) brings the handbook into the 21st century by treating the effects of new media (radio, television, computer) on forms of the language, as well as the longer established effects of newspapers. The second volume (edited by Alexander Bergs) then abstracts away from these details and moves outward to address theoretical concerns raised by the topics covered in Volume 1. Volume 2 first surveys resources for the studying and teaching of English (Section IX). Section X on interdisciplinarity (in particular literature and music) and historiography explores some of the debates involved in writing a history of English, questioning, for example, how the continuum of history is divided into accepted
Preface to Historical Linguistics of English
vii
“periods”, how oral and written forms of the language are accommodated in a history of English, and how new and perhaps “alternative histories” relate to the more established stories. This is followed by a history of the discipline of English historical linguistics itself, as it has developed in different parts of the world (Section XI). A significant part of Volume 2 covers changes in the English language as they have been theorized in various linguistic fields in the 20th century (Section XII). As Neogrammarian and Structuralist approaches are, to a great extent, embodied in the treatment of topics in Volume 1, this volume begins with later 20th century theories, including Generative Grammar, Construction Grammar, Lexical Functional Grammar, Rates of Change, Frequency, Lexical Diffusion, Grammaticalization, Lexicalization, and Language Acquisition. Related to the theoretical perspectives are new approaches which have been developed in the analysis of the history of English, including Historical Dialectology, Historical Sociolinguistics, Historical Pragmatics, Corpus Linguistics, Information Structuring, and Actuation/ Change from Below. Another important aspect of Volume 2 is its focus on the effects of language contact and the often neglected history of different varieties of English. It offers a section on language contact in the history of English, organized by contact languages, and supplemented by discussions of pidginization and creolization in the history of English and its varieties (Section XIII). Section XIV comprises historical sketches of more than ten varieties of English, and complementary theoretical discussions of dialect contact, diffusion, and supra-regionalization. The history of several second-language varieties is treated in Section XV, ending with a discussion of Global English. The beginning of a new millennium seems the right time for taking stock of the long span of scholarship in English historical linguistics and for surveying the field as a whole. Furthermore, the availability of electronic resources has changed the study of the history of English in fundamental ways, and it is important that a new handbook recognize this turning point in the study of English. Alexander Bergs, Osnabru¨ck (Germany) Laurel J. Brinton, Vancouver (Canada)
Contents
Volume I Acknowledgments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv In memoriam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xvii General abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xix
I Periods 1 2 3 4 5 6
Pre-Old English Jeannette K. Marsh . Old English Ferdinand von Mengden. Middle English Jeremy J. Smith . . . . . Early Modern English Arja Nurmi. . . Late Modern English Joan C. Beal . . Contemporary English Jim Miller . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. 1 19 32 48 63 78
7 Phonology Janet Grijzenhout . . . . . . . . . . . . . . . . . . 8 Prosody Donka Minkova . . . . . . . . . . . . . . . . . . . . . 9 Morphology Dieter Kastovsky . . . . . . . . . . . . . . . . . 10 Syntax Graeme Trousdale . . . . . . . . . . . . . . . . . . . . 11 Semantics and lexicon Elizabeth Closs Traugott. . . . 12 Idioms and fixed expressions Gabriele Knappe . . . . 13 Pragmatics and discourse Andreas H. Jucker . . . . . . 14 Onomastics Carole Hough . . . . . . . . . . . . . . . . . . . 15 Orthography Hanna Rutkowska . . . . . . . . . . . . . . . 16 Styles, registers, genres, text types Claudia Claridge.
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
97 113 129 148 164 177 197 212 224 237
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
255 272 294 313 325 340 362 373 385
II Linguistic Levels
III Old English 17 18 19 20 21 22 23 24 25
Phonology Robert Murray . . . . . . . . . . . . . . Morphology Ferdinand von Mengden . . . . . . Syntax Rafał Molencki . . . . . . . . . . . . . . . . . Semantics and lexicon Christian Kay . . . . . . . Pragmatics and discourse Ursula Lenker . . . . Dialects Hans Sauer and Gaby Waxenberger. Language contact Gernot R. Wieland . . . . . . Standardization Lucia Kornexl . . . . . . . . . . . Literary language Robert D. Fulk . . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
x
Contents
IV Middle English 26 27 28 29 30 31 32 33 34 35 36
Phonology Nikolaus Ritt . . . . . . . . . . . . . . . . . . . . Morphology Jerzy Wełna. . . . . . . . . . . . . . . . . . . . Syntax Jeremy J. Smith . . . . . . . . . . . . . . . . . . . . . Semantics and lexicon Louise Sylvester . . . . . . . . . Pragmatics and discourse Elizabeth Closs Traugott Dialects Keith Williamson. . . . . . . . . . . . . . . . . . . Language contact Herbert Schendl . . . . . . . . . . . . Standardization Ursula Schaefer . . . . . . . . . . . . . . Sociolinguistics Alexander Bergs . . . . . . . . . . . . . . Literary language Leslie K. Arnovick . . . . . . . . . . The language of Chaucer Simon Horobin . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
399 415 435 450 466 480 505 519 534 551 576
V
Early Modern English
37 38 39 40 41 42 43 44 45 46 47 48 49 50 51
Phonology Julia Schlu¨ter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Morphology Claire Cowie . . . . . . . . . . . . . . . . . . . . . . . . . . . . Syntax Elena Seoane. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lexicon and semantics Ian Lancashire. . . . . . . . . . . . . . . . . . . Pragmatics and discourse Dawn Archer. . . . . . . . . . . . . . . . . . Dialects Anneli Meurman-Solin . . . . . . . . . . . . . . . . . . . . . . . Language contact Laura Wright . . . . . . . . . . . . . . . . . . . . . . . Standardization Lilo Moessner . . . . . . . . . . . . . . . . . . . . . . . . Sociolinguistics Helena Raumolin-Brunberg . . . . . . . . . . . . . . Pronouns Ulrich Busse . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Periphrastic DO Anthony Warner . . . . . . . . . . . . . . . . . . . . . . The Great Vowel Shift Manfred Krug . . . . . . . . . . . . . . . . . . . Relativization Christine Johansson . . . . . . . . . . . . . . . . . . . . . Literary language Colette Moore . . . . . . . . . . . . . . . . . . . . . . The language of Shakespeare Ulrich Busse and Beatrix Busse .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
589 604 621 637 652 668 685 698 714 731 743 756 776 791 808
52 Phonology Charles Jones. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 Morphology Britta Mondorf . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 Syntax Bas Aarts, Marı´a Jose´ Lo´pez-Couso, and Bele´n Me´ndez-Naya . 55 Semantics and lexicon Marina Dossena . . . . . . . . . . . . . . . . . . . . . . 56 Pragmatics and discourse Diana M. Lewis . . . . . . . . . . . . . . . . . . . . 57 Dialects Susanne Wagner . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 Standardization Anita Auer. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 Sociolinguistics Erik Smitterberg . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
827 842 869 887 901 915 939 952
VI Late Modern English
VII 60 61 62
Standardization Prescriptive tradition Edward Finegan. . . . . . . . . . . . . . . . . . . . . . . . . . 967 The complaint tradition Tony Crowley . . . . . . . . . . . . . . . . . . . . . . . . . 980 Standards in the history of English Claudia Lange. . . . . . . . . . . . . . . . . 994
Contents
63 64 65 66
xi
Codifiers Carol Percy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . English language regard Dennis R. Preston and Jon Bakos . . . . Bible translations Thomas Kohnen . . . . . . . . . . . . . . . . . . . . . . Dictionaries and the standardization of English John Considine.
. . . .
. . . .
. . . .
. . . .
. . . .
1006 1020 1039 1050
. . . .
. . . .
. . . .
. . . .
. . . .
1063 1075 1089 1105
VIII English and the Media 67 68 69 70
Newspapers Udo Fries . . . . . Television Jane Stuart-Smith . Radio Ju¨rg Rainer Schwyter . Internet Theresa Heyd . . . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1119
xii
Contents
Volume 2 IX Resources 71 72 73 74 75 76
X 77 78 79 80 81
Early textual resources Kathryn A. Lowe Electronic/online resources Oliver M. Traxel Lexicographic resources Philip Durkin Teaching perspectives Michael Adams Textbooks Mary Blockley Online resources for teaching Beatrix Busse
Interdisciplinarity and Historiography Literature Andrew Johnston Music as a language – the history of an idea Nadja Hekal Periodization in the history of the English language Anne Curzan Myths of the English language; or, alternative histories of “English” Richard J. Watts Spoken and written English – orality and literacy Ursula Schaefer
XI History of English Historical Linguistics 82 83 84 85 86 87 88 89 90
XII 91 92 93 94 95 96 97 98 99 100 101 102
Overview Alexander Bergs and Laurel J. Brinton The historiography of the English language Jeremy J. Smith North America Thomas Cable Germany and the German-speaking countries Ilse Wischer The Netherlands and Belgium Peter Petre´ Northern Europe Risto Hiltunen East-Central and Eastern Europe Ire´n Hegedu˝s Southern Europe Teresa Fanego Asia Minoji Akimoto
New Perspectives, Theories and Methods Historical dialectology I. Keith Williamson Historical sociolinguistics Terttu Nevalainen Historical pragmatics Irma Taavitsainen Information structure and syntax in the history of English Bettelou Los and Ans van Kemenade The actuation problem revisited Richard J. Watts Corpus linguistics Merja Kyto¨ Frequency and language change K. Aaron Smith Lexical diffusion Betty S. Phillips Grammaticalization Lieselotte Brems and Sebastian Hoffmann Lexicalization Laurel J. Brinton Diachronic change and language acquisition Holger Diessel Generative approaches to English historical linguistics Bettelou Los
Contents
103 104
Construction Grammar Alexander Bergs Lexical Functional Grammar Cynthia L. Allen
XIII English in Contact 105 German and Dutch Jennifer Hendriks 106 French Janne Skaffari 107 Celtic and Celtic Englishes Markku Filppula and Juhani Klemola 108 Latin Letizia Vezzosi 109 Greek Brian D. Joseph 110 Norse Richard Dance 111 English in contact with “other” European languages Cristina Sua´rez-Go´mez 112 Native American Languages Keren Rice 113 Pidgins and creoles Suzanne Romaine 114 Middle English creolization David M. Trotter 115 African American English (AAE) early evidence Alexander Kautzsch
XIV Varieties of English 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131
Standard American English Richard W. Bailey Re-viewing the origins and history of African American Language Sonja L. Lanehart Regional varieties of American English Luanne von Schneidemesser Canadian English in real-time perspective Stefan Dollinger Standard British English Pam Peters Received Pronunciation Lynda Mugglestone Estuary English Ulrike Altendorf Regional varieties of British English Bernd Kortmann and Christian Langstrof Scots Robert McColl Millar English in Ireland Jeffrey L. Kallen English in Wales Colin H. Williams English in Australian/New Zealand English Marianne Hundt Cockney Sue Fox Diffusion David Britain Dialect contact Peter Trudgill Supraregionalization Raymond Hickey
XV Second-Language Varieties 132 133 134 135 136
English in India Devyani Sharma English in Africa Rajend Mesthrie Second-language varieties of English Daniel Schreier English-based Creoles Andrea Sand Global English Joachim Grzega
xiii
Acknowledgments
Foremost, the editors wish to thank the nearly 150 experts in English historical linguistics worldwide who contributed chapters, without whom these volumes would not exist. We are particularly grateful to those who wrote two chapters or who stepped in to fill gaps that arose late in the process of assembling the contributions. We would also like to thank our Advisory Board – Cynthia Allen, Merja Kyto¨, Donka Minkova, and Elizabeth Closs Traugott – who gave us invaluable advice in the initial stages of this project. Thanks too to Anne Curzan, who helped in the planning stages. Our student assistants provided invaluable assistance in the editing stage: Slade Stolar and Martin McCarvill of the University of British Columbia; Jens Bonk, Lisa Gratzke, Barbara Hagenbrock, Claudia Ko¨mmelt, Mona Matzke, Meike Pentrel, and Lena Probst of the University of Osnabru¨ck. At De Gruyter Mouton, we are grateful to former Publishing Director Anke Beck for inviting us to develop this project and to Uri Tadmour to seeing it to completions, to Barbara Karlson for her encouragement, patience, gentle prodding, and expert guidance, and to Ulrike Swientek for her production expertise. For her keen eye and soft touch in copy-editing, we are most appreciative of Catherine Every (of EveryWord), and for her meticulous indexing, we thank Vicki Low (of Scholar’s Cap). We extend our gratitude to all of the following scholars, who generously contributed their time and expertise in serving as referees for the articles contained in these volumes. Some went well beyond the call of duty and reviewed more than one article or both wrote and reviewed an article: Sylvia Adamson John Algeo Ulrike Altendorf Leslie K. Arnovick Richard Bailey Jo´hanna Barðdal Joan Bresnan Derek Britton Ulrich Busse Joan Bybee Deborah Cameron Ruth Carroll Jack Chambers Claudia Claridge Eve Clark Sandra Clarke Richard Coates Juan Camilo Conde-Silvestre John Considine Nikolas Coupland
Jonathan Culpeper Hubert Cuyckens Mary Catherine Davidson Hendrik De Smet Dagmar Deuber Hans-Ju¨rgen Diller Stefan Dollinger Bridget Drinka Edwin Duncan Stefan Evert Edward Finegan Olga Fischer Susan Fitzmaurice Robert Fulk Heinz Giegerich Eugene Green Peter Grund Trinidad Guzma´n Gonza´lez Martina Ha¨cker Antonette diPaolo Healey
xvi
Acknowledgments Lena Heine Juan Manuel Herna´ndez-Campoy Susan Herring Raymond Hickey Gary Holland Richard Ingham Matti Kilpio¨ John Kirk Marina Kolokonte Lucia Kornexl William Kretzschmar Barbara Kryk-Kastovsky Merja Kyto¨ Roger Lass Gerhard Leitner Christian Liebl Michael Linn Angelika Lutz T. W. Machan Michael K. C. MacMahon Christian Mair Murray McGillivray Daniel McIntyre Donka Minkova Marianne Mithun Rosamund Moon Bruce Moore Colette Moore Susanne Mu¨hleisen Pieter Muysken Robert Murray Minna Nevala
Hans Frede Nielsen Arja Nurmi Stephen B. Partridge Meike Pfaff Joanna Przedlacka Matti Rissanen Juhani Rudanko Mats Ryde´n Pingali Sailaja Joseph Salmons Holger Schmidt Anne Schro¨der Elena Seoane K. Aaron Smith Dieter Stein Merja Stenroos Patrick Studer Sali Tagliamonte Sanna-Kaisa Tanskanen Sarah G. Thomason Ingrid Tieken Boon von Ostade Carola Trips Uwe Vosberg Susanne Wagner Terry Walker Gregory Ward Brita Wa˚rvik John Wells Gernot R. Wieland Walt Wolfram Alison Wray Nuria Ya´n˜ez-Bouza
In memoriam
We commemorate those friends and colleagues who passed away since this project came into being. Without them, English historical linguistics will not be the same: Richard Bailey, Derek Britton, and Richard Hogg.
General abbreviations
ACC ACT ADJ ADV AN
Angl. AUX AP
C C COMPR DAT CP DEM
DM DU
EModE EWSax. FEM
Fr. GEN
Ger. Gk. Go. Grmc. IE IMP IND INF INFL INSTR
IP Kent. Lt. LModE LWSax. MASC
ME MED ModE NEG NEUT
accusative case active adjective adverb Anglo-Norman Anglian auxiliary adjective phrase consonant complementizer comparative dative case complementizer phrase demonstrative discourse marker dual Early Modern English Early West Saxon feminine French genitive case German Greek Gothic Germanic Indo-European imperative indicative infinitive inflected instrumental case inflection phrase Kentish Latin Late Modern English Late West Saxon masculine Middle English Middle English Dictionary Modern English negative neuter
xx
General abbreviations noun nominative case NP noun phrase O object OBJ objective case OE Old English OED Oxford English Dictionary OFr. Old French OFris. Old Frisian OHG Old High German ON Old Norse OSax. Old Saxon OV object-verb word order P person PASS passive PAST past PDE Present-day English PGrmc. Proto-Germanic PIE Proto-Indo-European PL plural PP prepositional phrase PREP preposition PRON pronoun PRTC participle PRES present PRET preterit S subject SG singular SUBJ subjunctive mood SUP superlative SOV subject-object-verb word order SV subject-verb word order SVO subject-verb-object word order SVX subject-verb-other parts of sentence word order T tense THM thematic vowel TMA tense-modality-aspect TVX topic-verb-other parts of sentence word order V verb V2 verb second V vowel VO verb-object word order VP verb phrase WGrmc. West Germanic WSax. West Saxon XP variable phrase XSV others parts of sentence-subject-verb word order N
NOM
General abbreviations XVS
> < Ø *
other parts of sentence-verb-subject word order changes to, becomes derives from no ending reconstructed form, ungrammatical form spelling
xxi
I Periods 1 Periods: Pre-Old English 1 2 3 4 5 6 7 8
Introduction Origins of English The phonology of Proto-Germanic Phonology: Proto-Germanic to Pre-Old English Morphology Syntax Summary References
Abstract The topic of this chapter is the pre-historic stage of English as it developed between the 5th-century Germanic migrations to Britain and its first attestations in the 7th century. The beginnings of Old English are situated with respect to the language’s closest West Germanic relatives as well as to its Indo-European linguistic heritage. The phonological system is traced from Indo-European through Proto-Germanic and West Germanic stages with a focus on those innovations that occurred during the pre-Old English period. Brief descriptions of Indo-European and Proto-Germanic morphological structure provide the basis of the sketch of pre-Old English morphology, while both phonological and morphological changes that later obscured these systems in the early development of English are illustrated. The chapter ends with a brief discussion of the development of pre-Old English syntax.
1 Introduction Though this chapter is titled “Pre-Old English”, there was, of course, no clear-cut division between the attested Old English (OE) language and what came before. What is meant by “Pre-Old English” here is the pre-historic stages of English, that is, the Germanic language spoken in Britain after the migrations of Germanic speakers (Germani) from their continental home, but prior to the language’s first textual transmission, i.e. from the 5th to the 7th century. The Northwest branch of Germanic from which English descends is only meagerly attested prior to and during this period in the form of runic inscriptions. We must therefore base our sketch of pre-Old English on comparative reconstruction of other Germanic and even Indo-European languages and then interpolate the specific features of this stage of the language using the first attestations of Old English. Thus the role of the Germanic linguistic inheritance on Old English will figure prominently in this chapter.
2 Origins of English The English language owes much of its character to its ancestry in the Indo-European (IE) family of languages. The Germani’s ultimate homeland is attested by a number of Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 1–18
2
I Periods classical sources including Caesar, Tacitus, and Jordanes, all of whom describe Germanic tribes living in northern Europe and along the North Sea coast. The Goths’ late 2nd century migration toward the Black Sea left the remaining northern and western branches of Germanic to develop separately. It was the tribes that remained along the shores of present-day Germany, southern Denmark, and the Netherlands after more southerly West Germanic tribes had pushed toward the Danube and the Alps that formed the linguistic stock of what would become Old English, Old Frisian (first attested from the 13th century), and probably some of Old Saxon as well (attested from the 9th century). This broad dialect group is referred to as North Sea Germanic or Ingvaeonic and the term “Anglo-Frisian” refers to the Ingvaeonic sub-grouping from which English, or at least dialects of it, derived. Archaeologists have observed a continuity of cultural artifacts between areas of Germanic settlement in Britain and those in the settlers’ original homelands on the continent and there are linguistic parallels as well which link Anglian, Kentish, and, to a lesser degree, Northumbrian dialects with Old Frisian (see Nielsen 1989: 53–65 for an overview of scholarship on these parallels). Generally speaking, this group of dialects was more innovative than the rest of West Germanic, likely due, at least in part, to the social upheaval and ensuing linguistic contact that was precipitated by the migrations and subsequent settlement of Britain. While the North Sea linguistic ancestry of the OE dialects is undisputed, there has been a more recent scholarly movement to trace some of the innovations seen in OE and Middle English (ME) texts to Celtic influence in Britain. Only about a dozen Celtic loanwords survive in Old English, these being mostly place names and names of geographical features. Traditional scholarship held that it was only under limited linguistic contact that Celtic could have had so little influence on the Grmc. dialects. But new scholarship suggests that Romanized Celts and Germanic people probably lived in close contact, sharing cultural items and communicating with each other in the languages of the invading Germanic tribes. Some scholars propose a contact situation in which the Celts, though far outnumbering the Germanic settlers, learned the language of the Germanic speakers imperfectly. The large ratio of bilingual Celts to invading Germanic speakers, in conjunction with the Celts’ limited access to Germanic, would have resulted in the Celts imposing a number of features of their native language onto their second language (pre-Old English) (see, for example, van Coetsem 1988: 7–45, 83–91; Thomason and Kaufman 1988: 35– 63; Guy 1990: 48–54, for discussions of the social and linguistic circumstances that foster imposition of first-language features onto the second language, instead of borrowing). The sheer number of partially bilingual Celts would have nearly ensured transmission of those features into the following generations of British Germanic speakers. For an introduction to the current scholarship in this area, see Flippula et al. (2002: 5–26). Many of the Old English grammatical handbooks treat the phonological and morphological development of Germanic from its IE ancestor. Among these are Luick (1964a [1914–21], 1964b [1929–40]), Wright and Wright (1925), Campbell (1959), Brunner (1965), Hogg (1992), and Hogg and Fulk (2011). More detail on the sound changes and morphological structures of the early Germanic stages are presented in Prokosch (1939), Krahe and Meid (1969), the essays in van Coetsem and Kufner (1972), and, more recently, in Ringe (2006). In the sections to follow we present an overview of both the features that English inherited from its Germanic ancestors and the changes
1 Periods: Pre-Old English
3
which occurred in the intervening periods that gave Old English its particular character. The phonology section consists of a description of the features which West Germanic inherited from its IE ancestor, followed by a description of the specific developments which occurred during the pre-OE period. Since changes in the phonology also had an impact on the morphology of the language, those changes will be introduced in the phonology section. The morphology section will provide an overview of the development of morphological categories and structure from the ancestors of Old English. Syntax, being considerably more difficult to reconstruct without substantial attestation, will comprise a final, brief, section of the chapter.
3 The phonology of Proto-Germanic 3.1 The consonant system The linguistic change that is most commonly used as a marker of the Germanic (Grmc.) language family is the First Germanic Consonant Shift – also referred to as Grimm’s Law – in which the entire system of IE stop consonants is alleged to have shifted. The version of the shift presented here is the traditional one and that most commonly assumed today. For details of an alternative reconstruction of the IE stop system and its ensuing shift into Germanic, the reader is referred to the Glottalic Theory proposed separately by Gamkrelidze and Ivanov (1995) and by Hopper (1973). The consonant system of late western Indo-European is traditionally reconstructed as having had the stops shown in Table 1.1. Table 1.1: Late western Indo-European stops
voiceless voiced voiced aspirates
labials
coronals
velars
labiovelars
p (b) bh
t d dh
k g gh
kw gw gw h
The First Germanic Consonant Shift shifted the IE voiceless stops, *p, *t, *k, *kw, to fricatives, f, θ, x, xw. The IE voiced stops, *b, *d, *g, *gw, then shifted into the vacated position of the voiceless stops, p, t, k, kw, and the voiced aspirates, *bh, *dh, *gh, *gwh, shifted to voiced fricatives, β, ð, ɣ, ɣ w. The consonants of later Latin loanwords did not undergo the same shifts as the native Grmc. consonants and therefore often demonstrate a more transparent reflex of the original IE stop. For example, the native Germanic development of the IE roots *de´km ‘ten’ and *bhra¯´ter ‘brother’ illustrate the Germanic consonant shift, while Latinate loanwords from these same roots do not show these shifts: IE *de´km > PGrmc.*texun (OE tı¯en) vs. Latin decimal; IE *bhra¯´ter > PGrmc. *bro¯´θar (OE bro¯þor) vs. fraternity (< Old French < Latin, in which IE *bh > Lt. f ). The accent of Indo-European was a pitch accent whose placement was morphologically and lexically determined. When the original IE pitch accent had preceded a medial voiceless stop, the stop spirantized to a voiceless fricative as predicted by Grimm’s Law, e.g. IE *bhra¯´ter > PGrmc. *bro¯´θar. But when a high pitch accent had followed the stop, a major exception to the expected outcome occurred. Presumably
4
I Periods due to the slack vocal folds and comparatively low pitch of the preceding unaccented syllable (D’Alquen 1988: 17–20; Page 1998: 186–188; Petrova 2004: 376–381), the fricative was perceived as voiced instead of voiceless as in IE *pate¯´r > PGrmc. *fa´ðe¯r. This exceptional voicing, known as Verner’s Law, affected all voiceless fricatives including */s/. Thus, IE *ge´us appears in OE ce¯osan ‘to choose’ (with /s/), but in the PGrmc. 1P ´ m, where the accent had followed the fricative, the /s/ was voiced to /z/ PL PRET *gusu and ultimately rhotacized to /r/ in Northwest Grmc. (cf. Section 4.1.5), thus OE curon ‘we chose’. The effects of Verner’s Law are evident in all of the Grmc. languages, though its appearance is much more restricted in Gothic. The resulting system is shown in Table 1.2. Following the application of Verner’s Law, the IE accent shifted to the root syllable. This increase in energy and duration of the root syllable would be responsible for enormous changes from the inception of the Grmc. languages through the Modern period. A number of these are described in the sections which follow. Table 1.2: The early Germanic consonant system
stops, voiceless fricatives voiceless voiced nasals liquids glides
labial
dental
alveolar
p f
θ
t s
k x
ð n l, r
ɣ
b m
palatal
j
velar
w (labiovelar)
In addition to the singleton consonants, geminate consonants could also appear in postvocalic environments. These developed in Proto-Germanic through contact assimilations of adjacent consonants and resulted in a system that included geminate versions of all of the stops, nasals, liquids, /s/, and probably of both of the glides as well. The voiced fricatives that developed from both IE voiced aspirated stops and from the voiced output of IE voiceless stops through Verner’s Law hardened into voiced stops (b, d, g, gw) at various times according to dialect and phonological environment. When following nasals, ß and ð probably became stops within the Grmc. period. The process would have continued in later periods with word-initial and perhaps post-liquid environments. The fricative articulation was preserved the longest for *ɣ, while *ð eventually developed a stop articulation in all environments in the West Grmc. branch of languages. Goblirsch (2003: 111–119) provides a detailed review of the scholarship on the development of the voiced fricatives in English and Frisian.
3.2 Indo-European to Proto-Germanic vowels The late IE vowel system consisted of long and short i, e, a, o and u. Non-syllabic high vowels could combine with preceding vowels to produce the diphthongs a¯̆ i, a¯̆ u, e¯̆ i, e¯̆ u, o¯̆ i, and o¯̆ u. Liquids and nasals could also function as syllabic nuclei in Indo-European but were reinterpreted by the early Germanic speakers as short u + resonant, e.g. IE *wṛg- > PGrmc. *wurk-. These reflexes are highly visible in the third and fourth principal parts
1 Periods: Pre-Old English
5
of Grmc. strong verb classes III and IV where u + resonant developed from the earlier syllabic resonant of the root syllable (Murray, Chapter 17: Section 2.4).
3.2.1 Long vowels Germanic preserved the distinction between long and short vowels from IndoEuropean with some shifting of the quality of those vowels within their respective systems. In the long vowel system Indo-European *a¯ moved up to merge with the existing *o¯ and IE *e¯ moved downward toward [æ¯] (also called “e¯1”). The vacated e¯ position was filled in Germanic by a monophthongization of IE *e¯i and by a front vowel with a relatively limited distribution that demonstrated *ı¯~*e¯ alternations in North Sea and North Germanic. The resulting e¯ is often referred to as e¯2. Original IE *ı¯, *o¯, and *u¯ remained as phonemes into Proto-Germanic. Original *ı¯ was reinforced by a monophthongization of IE *ei. Thus the PGrmc. long vowel system was as in Figure 1.1. ¯i
u¯
e¯ 2
o¯
¯ (e¯ 1) æ
Figure 1.1: Proto-Germanic long vowels
3.2.2 Short vowels In the short vowel system, movement was in the other direction so that IE *o and *a merged unconditionally into the existing *a, which continued into West Germanic. IE *e shifted to *i, and *u to *o in Proto-Germanic, but the application of these shifts was dependent on the following segments, demonstrating the developing preference for harmony between the stressed root vowel and the following vowels. Clusters of nasal + C preferred preceding high vowels, so nasal clusters facilitated the shift of *e to *i, but blocked the fall of *u to *o. We also see the effect of two kinds of distance vowel assimilations at this stage, a raising umlaut, “i-umlaut”, conditioned by a following high vowel or glide and a lowering umlaut (often called “a-umlaut”) conditioned by non-high back vowels. Thus the PGrmc. shift of *e to *i occurred unless an *a or *o followed in the next syllable. PGrmc. *u shifted to *o unless a nasal or *i followed and PGrmc. *i sometimes shifted to *e under similar conditions. The outcomes of some of these shifts are particularly evident in the principal parts of the OE strong class III verb. When a nasal follows the root vowel, as in PGrmc. *bendan-, *band, *bundun, *bundan, it is responsible for raising the *e of the present stem to i as well as for preventing the u of the past participle from being pulled to o by the a in the following syllable. Thus class III pre-OE principal parts bindan, band, bundun, bundan, but helpan (with no raising of e to i before l ), healp (cf. Section 4.2.3 for a discussion of the diphthongized vowel), hulpun, holpan (with lowered root vowel). The resulting system of short vowel phonemes appears in Figure 1.2.
6
I Periods i
u e a
Figure 1.2: Proto-Germanic short vowels
3.2.3 Diphthongs The Grmc. reflexes of the IE diphthongs suggest that the first element of the IE long diphthongs had generally shortened prior to subsequent Grmc. developments. Both IE *o¯i and *oi, for instance, developed into PGrmc. *ai. IE *e¯i and ei are exceptions to this pattern. As described above, *e¯i became e¯2 and *ei became PGrmc. *ı¯. IE *eu was retained into Proto-Germanic and was joined by a new diphthong, *iu. Thus, the PGrmc. diphthongs were *ai, *au, *eu, and *iu.
4 Phonology: Proto-Germanic to Pre-Old English The North Sea Germanic dialect which would develop into English and Frisian was differentiated from surrounding dialects by a number of phonological and morphological features. The consonantal features of the North Sea dialects that differentiated them from the rest of Germanic were the seeds of velar palatalization and a generalized loss of nasals before voiceless fricatives. The vowel system of this group also developed differently both through shifts in the quality of inherited vowels and in how vowels were affected by neighboring sounds. The most significant of these developments are outlined below.
4.1 Consonantal changes 4.1.1 Geminates The Grmc. inventory of geminate consonants was bolstered by the output of West Germanic consonant gemination, an innovation of the West Grmc. branch (with traces in North Germanic) that resulted from the effects of the resonants *j, *w, *l, and *r, on preceding consonants (except *r) following a short vowel, e.g. PGrmc. *lagjan > West Grmc. *laggjan; *wilja > *willja; *bitr- > *bittr-; *apl- > *appl-; *nakw- > *nakk-. As unstressed final vowels were reduced and lost, some originally medial geminates came to be word final. These tended to be simplified gradually in the early West Grmc. dialects. Degemination continues throughout the OE period where we find doublets with both geminate and simple final consonants, e.g., cynn ~ cyn, will ~ wil.
4.1.2 Palatalization Palatalization of velar consonants in the environment of adjacent (originally) front vowels is a feature shared by all of the North Sea Grmc. dialects, but whether the process began during an early period of their relative unity continues to be debated. If it did, then it is probable that pre-palatal *k developed an allophonically palatalized
1 Periods: Pre-Old English
7
articulation during this period and only later developed fully phonemicized assibilated phonemes in Old English, Old Frisian, and Old Saxon. (The reader is referred to the discussion of the OE palatalization of velars in Murray, Chapter 17: Section 3.3.)
4.1.3 The Pre-Old English consonant inventory PGrmc. ð was closed to d in all of West Germanic and β had become a stop in most positions by prehistoric English, while the fricative articulation of ɣ persisted intervocalically. Thus, the PGrmc series of voiced fricatives had become a series of voiced stops at this stage. The voiceless velar fricative had also begun to change. It had weakened to [h] word-initially and between sonorants and vowels, where it ultimately was lost. Its effect on preceding vowels (described in Section 4.2.3) suggests that it may also have been weakening in other coda positions during this period. The early English inventory had been enriched by West Germanic consonant gemination which produced geminates of all original (i.e. non-palatalized) stops, fricatives, nasals, and liquids except r. Geminate f appeared as in Old English and geminate glides usually combined with the preceding vowel to form sequences of diphthong + singleton glide. Thus the pre-OE system would have had the singleton consonants shown in Table 1.3. Table 1.3: Pre-Old English simple consonants labial stops voiceless voiced fricatives nasals liquids glides
p b f m
dental
alveolar
θ
t d s n l, r
palatal
velar k ɡ x
j
w (labiovelar)
4.1.4 Nasal loss and compensatory lengthening All of West Germanic underwent a loss of postvocalic nasals before PGrmc. *x (< IE *k). The North Sea dialects extended it to apply to postvocalic nasals before any voiceless fricative. The nasal cluster had the usual raising effect on preceding vowels, but the loss of the nasal conditioned a compensatory lengthening of the preceding short vowel. In Old English the nasalized, now lengthened, a¯˜ appears as o¯. Thus we have from late PGrmc. *sanft, gans, kunþs > OE so¯ft, go¯s, and cu¯þ with nasal loss, compensatory lengthening, and raising of the original a to o¯. This change is responsible for the now opaque relationship between PDE bring and brought, the latter having gone through the intermediary stages PGrmc. *branx-te > West Grmc. *bra¯˜ xte > OE bro¯hte.
4.1.5 Rhotics and their effects West Grmc. languages are also marked by a rhotacism of the IE *s that had undergone voicing to /z/ as a result of Verner’s Law. Proto-Germanic already had a rhotic which
8
I Periods was presumably coronal in articulation (see Denton 2003: 15–16, 19–30 for a discussion of the articulatory qualities of early Grmc. rhotics and their articulatory effects in OE dialects). The rhotacism of z eventually led to a merger with the original r which is visible in the third and fourth principal parts of strong verbs that had originally had medial s voiced through Verner’s Law (see examples in Section 3.1). But in word-final or unstressed position, rhotacized z was lost. This loss makes for a difference between the 1/2P SG personal pronouns in North Sea Germanic and those in the rest of West Germanic: OE DAT SG me, þe compared to OHG mir and dir.
4.2 The vowel system By the end of the Grmc. period, the short vowel system had only one low vowel phoneme and the mid back vowel only existed as an umlauted allophone of /u/ before mid and low vowels. Following this period the short vowels were further modified by their phonological environments and the long vowel system was enriched by the monophthongization of the PGrmc. diphthongs. By the start of the OE period PGrmc. *ai had become long a¯ and *au had become long æ¯a, written . Reflexes of PGrmc. *eu and *iu (long or short) remained largely distinct in Mercian and Northumbrian dialects (appearing as eo and io, respectively), but had merged in most environments in the earliest West Saxon (WSax.) texts, though both spellings remained. The long high and mid vowels (ı¯, e¯2, u¯, and o¯) continued into the Pre-OE period, but e¯1 underwent different developments in the various OE dialects, appearing as æ¯ in West Saxon and as e¯ elsewhere. When followed by a nasal, it appeared as o¯ in all of Old English. The WSax. æ¯ was irregularly retracted before mid and high back vowels in the following syllable (referred to as u-umlaut), but otherwise continued as a front vowel.
4.2.1 Anglo-Frisian brightening and retraction A similar process of allophonic split before nasals occurred with PGrmc. short *a (< IE *a and *o). While it shows up as a in the rest of West Germanic, in the Anglo-Frisian area it originally developed two allophones, a back variant before nasals and a front variant everywhere else. The back variant is alternately spelled either or in Old English as in the doublet mann ~ monn. Since this nasalized vowel was distinct from the existing o and eventually merged with a in most of Old English, it probably had a quality similar to [ɔ]. The non-nasalized variant, [æ], is the output of the process called “Anglo-Frisian brightening” (AFB) or “First Fronting” in the entire Ingvaeonic area. This variant behaved as a true front vowel, diphthongizing in Old English breaking environments (cf. Section 4.2.3) and palatalizing some velar consonants, e.g. PGrmc. gastiz > (AFB) gæste > (palatalization) g˙easte > (i-umlaut, cf. Section 4.2.4) g˙ieste. A subsequent process of retraction, however, pulled the front vowel back to a before an immediately following w in all dialects, while Anglian also retracted the front vowel before rC, and Northumbrian before lC.
4.2.2 Restoration of a The result of Anglo-Frisian brightening was subject to the early mutating effects of back vowels in the following syllable. This process, known as “restoration of a”, foreshadows
1 Periods: Pre-Old English the back vowel umlauts of the early OE period and is visible in the OE masculine and neuter a-stem paradigms. When the root vowel of the nominative/accusative, genitive, and dative singular is æ, as in dæg˙ ‘day’, dæg˙es, dæg˙e, it appears as a in the plural due to the retracting effect of the suffixes’ back vowels: NOM/ACC dagas, GEN daga, DAT dagum. The phonological conditioning and particular interactions of Anglo-Frisian brightening, retraction, restoration of a, and breaking are responsible for much of the dialectal variation in Old English and are discussed in detail, in the modern Old English grammatical handbooks (e.g. Brunner 1965: 38–46, 54–60; Luick 1964a [1914–21]: 122– 166; Campbell 1959: 50–64; Hogg 1992: 76–101; Lass and Anderson 1975: 59–69; Lass 1994: 39–44; Wright and Wright 1925: 38–68).
4.2.3 Breaking Breaking was a process of front vowel mutations conditioned by the following consonantism. Though preserved in its most regular form in the WSax. dialect of Old English, breaking appears to have applied with slight variation in all of pre-Old English and was even shared, in part, by Old Frisian and Old Norse. The graphic realization of the output of breaking in Old English was , , and from the high, mid, and low front vowels, respectively. Prior to breaking, the Grmc. languages had no short diphthongs which are typologically marked. Whether those sounds represented by short digraphs in Old English were true short diphthongs phonemically is part of a larger controversy, though OE metrics confirm that they were indeed distinguished from long diphthongs quantitatively. Stockwell and Barritt (1951: 14) first questioned the literal reading of these digraphs and many scholars since have argued that the output of breaking was monophthongal and that the addition of a central or back vowel grapheme indicated either a retraction of the original vowel quality or a secondary articulation on the following consonant. The entire controversy is nicely presented in brief in Lass and Anderson (1975: 75–83) (see also Murray, Chapter 17: Section 3.2), while more recently, White (2004: 58–59) takes up Daunt’s (1939) argument that the OE spelling pattern had Irish origins. Breaking appears to have applied only in stressed syllables in the pre-OE period and was more regular the lower and the shorter the vowel. Early textual evidence suggests that the breaking environments originally caused the front vowels to develop central- or back-vowel off-glides at the corresponding height, /i/ > [iu], /e/ > [eo], /æ/ > [æa]. The factors which conditioned breaking were the reflex of PGrmc. /x/, whether alone or followed by another consonant, r + C, w, and, to a lesser degree, l + C. The conditioning factors of breaking are generally assumed to have had back articulations that spurred a transitional glide between the palatal front vowel and the following consonant. Thus x and w are velar, the l of lC clusters are presumed velarized, and the r in rC clusters were also presumed to have had a back articulation similar, perhaps, to the Scots “burred” r. Howell (1991: 83–105) has refuted this assumption, demonstrating that the kinds of diphthongizations seen in breaking rarely occur with sequences of stressed front vowel + velar liquid or x in modern Grmc. dialects, though analogous diphthongizations are quite common when the coda contains a weakened, non-velar, articulation of these same segments. Howell adduces dialectal evidence that all breaking factors had less constricted, even vocalic, articulations in coda
9
10
I Periods positions and Denton (2003: 21–30) makes a similar argument about OE r on articulatory grounds. The preserved effects of breaking with limited retraction and restoration of a in West Saxon contribute to the marked difference in sound and appearance of this dialect compared to that of Anglian and Northumbrian. This fact was compounded by Anglian’s later monophthongization or “smoothing” of many of these diphthongs. The output of breaking has morphological consequences, perhaps the most obvious being the subclass of class III strong verbs which are characterized by a short root vowel followed by a liquid-consonant cluster, e.g., Pre-OE. *werþan, *wærþ; *helpan, *hælp > OE weorþan, wearþ; helpan, h(e)alp. Breaking before h ( y, o > œ, a > æ, though it also raised the low and mid front vowels with less regularity. (See Section 5.1 in Murray, Chapter 17 as well as the OE grammatical handbooks for a description of the conditioning and output of i-umlaut in Old English). The effects of i-umlaut are responsible for the root vowel change in the mutated plurals: NOM SG mann, go¯s, mu¯s; NOM PL menn, gœ̅ s ~ ge¯s, mȳ s < earlier *mann-i, go¯s-i, mu¯s-i Even more visible are the mutated vowels in the 2/3P PRES IND of strong verbs which originally had the suffixes -is and -iþ, respectively: c˙e¯osan 2/3P SG PRES IND c˙¯ıest < *ce¯osis, *ce¯osiþ; faran, færst < *farist, færþ < *fariþ. The entire system of class I weak verbs is subject to i-umlaut from the j of the -jan suffix that had originally marked the class. This fact is only evident when compared to the unumlauted root from which a class I weak verb was derived, e.g., early OE dœ̅ man ‘to judge’ < early West Grmc. *do¯m-jan-.
4.2.5 Stress and its immediate effects A dominating stress accent on the root syllable was one of the hallmarks of the Grmc. languages, but a secondary stress accent on polysyllabic words was also likely, though scholars don’t agree on the rules of its placement. What we can surmise about the placement of all stresses in early Germanic comes from the metrics of the first attested languages and from the manner and order in which medial and final syllables were reduced in the period prior to the first transmission of texts. This reduction began in word-final position with the loss of short unstressed vowels. Final unstressed diphthongs were monophthongized and unstressed medial vowels in open syllables were often lost prior to the earliest OE texts. Final nasals were also lost eventually, leaving the nuclear vowels exposed to a new round of reductions. Generally more resistant to loss were unstressed high vowels. Their ultimate loss in both unstressed final and medial syllables
1 Periods: Pre-Old English
11
was usually conditioned by a preceding heavy foot, i.e. a single stressed heavy syllable (CV̅ , CVV, or CVC(C)) or a stressed light syllable followed by another syllable. This conditioned loss was responsible for -u~-Ø and -i~-Ø alternations in the West Grmc. u- and i-stem nominal paradigms as illustrated by the OE examples in Table 1.4. The nominative singular of feminine o¯-stems also shows an alternation between -u after light stems and -Ø after heavy, e.g. giefu vs. la¯r. Table 1.4: High vowel loss in nominal paradigms i-stem, SG NOM ACC GEN DAT PL NOM/ACC GEN DAT
u-stem,
MASC
MASC
light
heavy
light
heavy
win-e (< -i) win-e (< -i) win-es win-e win-e, -as win-e, -as win-um
giest (< -i) giest (< -i) giest-es giest-e giest-as giest-as giest-um
sun-u sun-u sun-a sun-a sun-a sun-a sun-um
feld (< -u) feld (< -u) feld-a feld-a feld-a feld-a feld-um
While unstressed vowels tended toward reduction, the stressed root vowels of the North Sea Grmc. and pre-OE periods were affected in a different manner, being further differentiated through the assimilatory effects of following sounds. Sound changes of this type included breaking, retraction, and the umlauts described above.
5 Morphology 5.1 Morphological structure Pre-Old English inherited a large percentage of its word stock from Indo-European, though at least 30% of its lexicon may have come from other sources. Vennemann (2003: xiii–20) provides an introduction to some of the issues inherent in identifying IE origins for both Germanic lexemes and morphosyntactic structures while he provides background for innovative theories of early Germanic contact with non-IE languages. Germanic did inherit much of its morphological structure from its IE ancestor along with its primary word-formation processes of compounding and derivation. IE roots were monosyllabic CVC structures with slots for a nasal, liquid, or glide on either side of the vowel. While the consonants of the root remained fairly stable across morphological categories, the radical vowel could vary between the e- or “full” grade, the o-grade, and the zero-grade, according to the word’s function. This kind of vowel alternation is the source of many of the ablaut, or root vowel substitution, patterns in the Grmc. languages. Various types of suffixes could be added to the root and these also could display different vowel grades within the same consonantal context. The position of the IE accent was partially dependent on the grade of the root and of its suffix. Thus, the overall structure of all IE morphology was that of root + suffix + inflectional ending, the combination of root + suffix constituting the stem. This structure continued to be the foundation of the nominal, adjectival, and verbal classes of the Grmc. languages.
12
I Periods
5.2 Nominal morphology 5.2.1 The case system The IE nominal categories were of three types: the nouns and adjectives, the demonstrative pronouns, and the personal pronouns. All of these were marked for case and number. Indo-European probably had eight cases which indicated the relationship of the noun phrases in the sentence both to each other and to the action of the verb. These cases were the nominative, accusative, dative, ablative, locative, instrumental, and vocative. Four cases were preserved in West Germanic with a vestige of a fifth. The nominative remained the case of the subject of the sentence (as well as of predicate nominals), but was also used for direct address. The accusative remained for the direct object, duration of time, and extent of space. The dative expressed a less direct impact or reference of the action of the verb on a noun such as with the indirect object, motiontoward, and many of the functions of the original locative and often the instrumental. The genitive was used for possession, for a part of a larger whole, and with certain adjectives and prepositions. In Old English and Old Frisian the instrumental forms were only preserved in the demonstrative and interrogative pronouns and in the strong adjective declension.
5.2.2 Gender Late Indo-European had three genders which were preserved into the OE period: masculine, feminine, and neuter. While these categories probably had some connection to real-world physical and/or cultural characteristics at one time, by early Germanic they largely served only a grammatical function. Masculine and neuter nouns, pronouns, and adjectives are closely related and always share a number of endings. The difference between the two genders often lies in the way in which the nominative and accusative singular are marked. While these categories may be formally differentiated in the masculine, they are always identical in the neuter singular, a feature that goes back to Indo-European (e.g. NEUT NOM and ACC SG hit ‘it’, scip ‘ship’). In the most common declensions the IE feminine differed from the masculine and neuter in the presence of the feminine’s *a¯ theme vowel, which developed into PGrmc. *o¯. Forms of other feminine declensions were also heavily influenced by the *o¯-stem endings. Though in the earliest OE paradigms the endings of the feminine are wholly different from those of the masculine and neuter, the endings of all three genders of all stem types derived from a single set of inflectional endings.
5.2.3 Nouns and adjectives Nouns and adjectives were indistinguishable in Indo-European and continued to share many of their endings into the OE period. It was the particular quality of the theme or stem vowel (or the lack thereof) which determined a noun’s or adjective’s class. Each class had a particular set of inflectional endings associated with it which likely derived from a single set of endings for all noun classes in Indo-European. The differences in inflectional endings that we see across the early Grmc. paradigms are the result of a combination of factors among which are
1 Periods: Pre-Old English
13
(1) variations in the original placement of the IE accent which resulted in different grades of the stem vowel and different developments of IE *s (which could appear as either OE s or r, depending on the application of Verner’s Law) (cf. Section 3.1); (2) the coalescence of stem vowels with the endings; and (3) the Grmc. reductions of final syllables. An example of the masculine a-stems, feminine o-stems and masculine and feminine n-stems illustrate this development below. The reconstructions have been simplified somewhat for clarity in Table 1.5. Table 1.5: Reflexes of the Proto-Germanic case inflections in Old English declensions PGrmc.
a-stem
o-stem
i-stem
u-stem
n-stem
SG
MASC
MASC
FEM
MASC
MASC
MASC
NOM ACC
-V-z -V-m
stan < az stan < am
gief-u < o¯´ gief-e < o¯´n
win-e < iz win-e < im
sun-u < uz sun-u < um
GEN
-V-so
stan-es < a´so
gief-e < o¯z
win-es < iza
sun-a < auz
DAT
-V-i
stane- < ai
gief-e < ai
win-e < ı¯
sun-a < au
nam-a < o¯n nam-an < anam nam-an < in(e/a)z nam-an < ini
NOM
-V-z(ez)
gief-e < o¯z
win-e < ı¯z
sun-a < iuiz
nam-an < anez
ACC
-V-nz
gief-e < o¯´nz
-V-n
gief-e < o¯n
win-e < (NOM PL) win-a < io¯n
sun-a < uns
GEN
stan-as < o¯z(ez) stan-as < (NOM PL) stan-a < o¯n
DAT
-V-miz
stan-um < amiz
gief-um < o¯miz
win-um < imiz
sun-um < umiz
nam-an < anunz nam-an < anon nam-um < anmiz
PL
sun-a < o¯n
The classes that became dominant in Germanic were the a-stems, o¯-stems, and the weak n-stems. The Grmc. i- and u-stems, though still viable, were no longer robust since many of their former members had moved over to a- and o¯-stem declensions. Other minor classes could be marked by reflexes of a consonantal suffix added to the root or by a lack of theme as in the athematic or root nouns, the latter marked in Old English by i-umlauted root vowels. (See von Mengden, Chapter 18: Section 1.2.8 for a description of these.) The adjectives originally had the same thematic classes as the nouns, including a- and o¯-stems (as well as ja- and jo¯-stem subtypes), i-stem, and u-stems, but these were heavily influenced by the pronominal declension in the transition from Indo-European to Germanic. The resulting strong adjectival endings are consequently a mixture of the strong nominal and the demonstrative pronominal declensions. By the early OE period most adjectives of the minor declensions, as well as many nouns of the minor declensions, had moved over to those of the a- and o¯-stems and the ja- and jo¯-stems. A major innovation of the Grmc. languages was the development of a weak adjectival declension similar in form to the weak n-stems nouns. The weak adjectival suffix derived from IE *-en-/-on- which appears in Latin and Greek nicknames Cato (GEN Catonis) ‘smart/shrewd (one)’ < catus ‘smart, shrewd’ and Greek Strabo¯n ‘squint-eyed (one)’ < strabos ‘squint-eyed’. This suffix was probably used in Grmc. phrases like the precursor of OE se blinda mann to mean ‘the blind one, a man’. The individualizing character of this suffix was gradually grammaticalized into the form of the adjective that
14
I Periods appeared with definite noun phrases, while the strong adjectival suffixes came to be used with indefinite noun phrases (Jasanoff 2008: 205; Krause 1968: 175; Osthoff 1876: 120–133). All nominal elements of the noun phrase agreed in case, number, and gender. Comparative and superlative adjectives were formed from suffixation to the adjectival stem. Two West Grmc. suffixes, *-iz- and *-o¯z-, were responsible for the OE comparatives which both developed into OE -r- via Verner’s Law and subsequent rhotacism (cf. Section 3.1 and Section 4.1.5). The original -i- of the first of these caused umlaut of the preceding vowel. The OE superlatives -est and -ost/-ast also developed from two different suffixes, *-ist- and *-o¯st-, respectively, the first of which also caused i-umlaut of the preceding vowel. Thus Old English has root vowel alternations in some adjectives like eald/ieldra/ieldest and long/lengra/lengest, but not in others earm/earmra/ earmost. Adverbs were also derived from adjectives. (See von Mengden, Chapter 18: Section 1.3.4 for an overview of these and for a description of numerals.)
5.2.4 Pronouns Old English had two demonstrative pronouns and an interrogative pronoun that agreed with the other members of their noun phrase in case, number, and gender, though there was no gender distinction in the plural. These three pronominal paradigms derive from demonstrative pronouns in the earlier IE language and show substantial similarities in their endings. The main demonstrative was the se/þæt/se¯o ‘the, that (one)’ paradigm which served as both definite article and the unmarked demonstrative pronoun. Its suppletive merging of the IE *s- and *t- pronominal bases in a single paradigm was evident as far back as Greek and Sanskrit. A second, derivative, pronoun came to serve as the proximal demonstrative, þes/þis/þe¯os ‘this (one)’, which derived from the same IE *t- base as the other demonstrative, but with the addition of an *-s(s)- suffix following the vowel. This pronoun could also function both pronominally and as a determiner. Only Northwest Germanic has a paradigm of this particular construction. Both demonstratives preserved a fifth case form in the masculine and neuter singulars only. This was þy¯~þon in the se/þæt/se¯o paradigm and þy¯s in the þes/þis/þe¯os paradigm. These forms, though labeled “instrumental” in the grammars and handbooks, neither derive directly from an earlier instrumental case nor are they used only to express instrumentality. Rather they are used for adverbial and idiomatic expressions such as þy¯ geare ‘in that year’, þy¯ læ¯s ‘lest’, æ¯r þon ‘before that’, þy¯ ma¯re þy¯ … ‘the more, the…’ The Grmc. interrogatives have clear cognates in other IE languages, descending from IE roots in *kw-. Though the inflectional endings of this paradigm are quite similar to those of the demonstrative pronouns, the interrogatives differ in that they have no plural forms and the singular forms combine the masculine and feminine into a single, animate, category. Outside of the nominative in hwa¯, the endings of the masculine/feminine forms resemble those of the demonstratives. The neuter has a separate nominative/accusative singular form, but is otherwise identical to the masculine-feminine paradigm. The Grmc. first and second personal pronouns are also derived from IE material. Though Gothic demonstrates that the dual pronouns originally required agreement with dual verbal forms in early Germanic, by Old English the dual forms of verbs had been lost and dual pronouns therefore agreed only with plural verbs. Unlike the
1 Periods: Pre-Old English first and second person pronouns, the Grmc. third person pronouns are marked for gender as well as for case and number. This fact may be due to their origins in demonstrative bases that were, themselves, marked for gender. Indeed, the personal pronouns’ heritage is heterogeneous. Four separate bases fed into the development of the various Grmc. third person pronominal paradigms. Those responsible for the English system were first the IE deictic in *k- (PGrmc. *x), which developed into the OE singular forms in h-, i.e. he, hine, heo, his, etc. The second base derived from an IE demonstrative pronoun in *ei- ~ -i- that formed the base of the OE plurals in h- (the initial h- was probably added later by analogy to the singular forms in h-) (Lass 1994: 141). ProtoGermanic had no common pronoun for introducing relative clauses, but the indeclinable pronoun þe developed as the most common means of expressing this kind of syntactic relationship in Old English.
5.3 Verbal morphology The ablauting verbs of Indo-European are divided into seven classes of strong verbs in Old English, all of which employ ablaut in conjunction with suffixation to express differences in tense, mood, and number. A small group of anomalous verbs also derived from IE origins. Perhaps the most significant innovation in the Grmc. verbs was the shift of what was primarily an aspectual system to a binary one of present versus past tense. As with the nouns, the dual was lost from the verbs. The indicative was preserved as the mood of declaratives and the second person imperatives continued to express commands, but only in the present tense. The infinitive was a present tense verbal noun which retained case marking in the early Grmc. inflected infinitive forms of which Old English preserves the original dative case. The present participle was a verbal adjective whose form is cognate with that of other IE languages. A Grmc. subjunctive, used in both present and preterit tenses, was formed from the IE optative, which had expressed the wish of the speaker. In the Ingvaeonic languages the three persons of the indicative plural were collapsed into one and the subjunctive had only a single form in each of the singular and plural of both tenses. West Germanic retained only the active voice with vestiges of the passive. The passive/past participle was a passive adjective when transitive, but was simply preterit when intransitive. OE ha¯tte/ha¯tton ‘is or was/ are or were called’ preserves the original passive, while all other uses of a passive in Old English were periphrastic constructions using weorþan or be¯on/wesan with the passive participle.
5.3.1 Strong verbs In the West Grmc. languages each strong verb had four principal parts: a present stem, a preterit first and third person singular stem, a preterit plural and second person singular stem, and a passive participle stem. For at least the first five classes it is presumed that there was root stress on the first two principal parts and suffix stress on the third and fourth. This assumption is supported by the output of Verner’s Law visible in the alternation between root-final voiceless fricatives in the first two principal parts and voiced fricatives or stops in the third and fourth (Section 3.1). For a detailed discussion of the development of the PGrmc. strong verbs from both Indo-European and non-IE origins, the reader is referred to Mailhammer (2007).
15
16
I Periods
5.3.2 Weak verbs The weak verbs, though an innovation of the Grmc. languages, formed their stems from the same morphological material as the strong verbs, i.e. a root – often in the o-grade – followed by a stressed suffix. What identified the weak verbs was a dental suffix marking the preterit rather than a root vowel alternation. One of the more popular hypotheses regarding the origin of the dental suffix was that it was a grammaticalization of the preterit forms of the verb to do suffixed to the verbal stem. The formal similarity between the Gothic forms of the weak preterit suffix and those of the PGrmc. verb *don constitute the basis for this claim, e.g. PGrmc. 3P PL *da¯dun/dedun ‘they did’ compared to the Go. 3P PL PRET IND suffix -de¯dun. Germanic probably had four classes of weak verbs all of which contained verbs derived from other categories. The first class suffix was from IE *-eja- which became *-ja in Germanic. The palatal glide was responsible for both i-umlaut of the root vowel and for gemination of the preceding consonant (cf. Section 4.2.4 and Section 4.1.1, respectively) when the j was preserved into West Germanic. In the second and third singular present indicative, the imperative, and the entirety of the preterit system, the original glide merged with following vowels to form a high front vowel rather than a glide. Because i did not spur gemination of a preceding consonant as j did, we find no gemination in these forms of the verb, though i-umlaut is present in all forms of the weak class i. In the second class the suffix derived from IE *-a¯je-/-a¯jo- which developed to *-o¯i- in Germanic. At this stage there was no longer a trigger for gemination of the root consonant since the original *j had collapsed into a diphthong with the preceding vowel. Unlike in class one, there is therefore no gemination in class two and also no i-umlaut in the root. If the following inflection began with a vowel, then the thematic -o¯i- sequence generally became -ia- in Old English as in the infinitive PGrmc.*lufa¯je-onom > OE lufian. If the suffix were followed by a consonant, the glide of the diphthong was lost and the long, unstressed, o¯ developed regularly to a. Thus we find 2/3P SG PRES IND lufast, lufaþ in class two where we found fremest, fremeþ in class one. OE class three verbs have no reflex of a stem vowel, though one does appear in this class elsewhere in Germanic. Thus, the OE personal endings are added directly to the root. There is no trace of a fourth class of weak verbs in Old English and even the third class has been reduced through the migration of its verbs into the first two classes.
5.3.3 Preterit-present verbs The preterit-present verbs constitute a third system of verbs in the Grmc. languages. They began as strong verbs whose past tense forms developed stative, present-tense, meaning. The strong class I preterit singular, wa¯t < *wı¯tan ‘to see’ (cf. Lt. vı¯dere), for instance, came to mean ‘I know’ (presumably through the development ‘I saw’ > ‘I saw, therefore I know’). In order to express a past tense of this new meaning, speakers created a new preterit form using a dental suffix which has sometimes been associated with the dental preterit marker of the weak verbs, though there are other theories of its origin (see, for example, Prokosch 1927: 334–335). Thus the preterit-presents have a present tense which resembles a strong preterit and a new weak preterit tense. It preserves some of the more archaic endings of the IE perfect seen in the present singular forms in which the earlier -e of the 1/3P SG was lost in West Grmc., e.g., sceal, and in
1 Periods: Pre-Old English the -t of the 2P SG which predated the -st that predominates elsewhere, e.g., scealt. The preterit marker was added to the original perfect stem with no intervening stem vowel: wiste/wisse and wiston (< *wit-te and *wit-ton, respectively). Old English had preterit-present verbs for each of the first six classes of strong verbs.
6 Syntax Though case marking on noun phrases allowed early Grmc. word order considerable flexibility, the unmarked word order for main clauses was OV, an order inherited from Indo-European. Modifiers most commonly occurred after the phrasal heads. Prepositions far outnumbered postpositions and were closely related to verbal particles which also preceded their heads. In Pre-Old English and later these particles were often written as separate words, although by the OE period we consider them verbal prefixes. Between the Northwest Grmc. runic inscriptions of the 3rd-7th centuries and the first OE texts, SVO word order became somewhat more common and modifiers also commonly appeared before their heads (Lass 1994: 218–222). Pre-Old English must have experienced substantial variation in word order as the language shifted from OV, which dominated the Northwest Grmc. runic inscriptions, to allowing verb-second order, which was quite common in Old English (Lass 1994: 225).
7 Summary The history of the Germanic speakers who migrated to Britain is characterized by migration, ongoing social upheaval, and heavy linguistic contact. It is therefore not surprising that the Pre-OE period was one of substantial linguistic change. Indeed, Old English may be the most innovative of the early Grmc. languages in terms of sound change alone. Though social instability and linguistic contact continued to spur innovation in the history of English, the process of rapid differentiation began with the first waves of migration to Britain and can be seen in the first OE texts.
8 References Brunner, Karl. 1965. Altenglische Grammatik. Nach der Angelsa¨chsischen Grammatik von Eduard Sievers. Dritte, neubearbeitete Auflage. Tu¨bingen: Niemeyer. Campbell, Alistair. 1959. Old English Grammar. Oxford: Oxford University Press. van Coetsem, Frans. 1988. Loan Phonology and the Two Transfer Types in Language Contact. Dordrecht/Providence, RI: Foris. van Coetsem, Frans and Herbert L. Kufner. 1972. Toward a Grammar of Proto-Germanic. Tu¨bingen: Max Niemeyer. Curzan, Anne and Kimberly Emmons (eds.). 2004. Studies in the History of the English Language II: Unfolding Conversations. Berlin/New York: Mouton de Gruyter. D’Alquen, Richard. 1988. Germanic Accent, Grammatical Change and the Laws of Unaccented Syllables. New York: Peter Lang. Daunt, Marjorie. 1939. Old English sound changes reconsidered in relation to scribal tradition and practice. Transactions of the Philological Society 1939: 108–137. Denton, Jeannette M. 2003. Reconstructing Germanic *r. Diachronica 20(1): 11–43. Flippula, Markku, Juhani Klemola, and Heli Pitka¨nen. 2002. Introduction: Early contacts between English and the Celtic languages. In: Markku Flippula, Juhani Klemola, and Heli Pitka¨nen (eds.), The Celtic Roots of English, 1–26. Joensuu: University of Joensuu, Faculty of the Humanities.
17
18
I Periods Gamkrelidze, Thomas V. and Vjacheslav V. Ivanov. 1995. Indo-European and the Indo-Europeans. Trans. by Johanna Nichols. 2 vols. Berlin/New York: Mouton de Gruyter. Goblirsch, Kurt Gustav. 2003. The voicing of fricatives in West Germanic and the partial consonant shift. Folia Linguistica Historica 24(1–2): 111–152. Guy, Gregory. 1990. The sociolinguistic types of language change. Diachronica 7: 47–67. Hogg, Richard M. 1992. A Grammar of Old English. Vol. I: Phonology. Oxford/Cambridge, MA: Blackwell. Hogg, Richard M. and R. D. Fulk. 2011. A Grammar of Old English. Vol. II: Morphology. Oxford/ Cambridge, MA: Wiley-Blackwell. Hopper, Paul J. 1973. Glottalized and murmured occlusives in Indo-European. Glossa 7(2): 141–166. Howell, Robert B. 1991. Old English Breaking and its Germanic Analogues. Tu¨bingen: Max Niemeyer. Jasanoff, Jay. 2008. Gothic. In: Roger D. Woodard (ed.), The Ancient Languages of Europe, 189– 214. Cambridge: Cambridge University Press. Krahe, H. and W. Meid. 1969. Germanische Sprachwissenschaft. 7th edn. Berlin: Walter de Gruyter. Krause, Wolfgang. 1968. Handbuch des Gotischen. 3rd edn. Munich: Beck. Lass, Roger. 1994. Old English: A Historical Linguistic Companion. Cambridge: Cambridge University Press. Lass, Roger and John M. Anderson. 1975. Old English Phonology. Cambridge/New York: Cambridge University Press. Luick, Karl. 1964a [1914–21]. Historische Grammatik der englischen Sprache, Vol. I, Part 1. Stuttgart: Bernhard Tauchnitz. Luick, Karl. 1964b [1929–40]. Historische Grammatik der englischen Sprache, Vol. I, Part 2. Stuttgart: Bernhard Tauchnitz. Mailhammer, Robert. 2007. The Germanic Strong Verbs. Berlin/New York: Mouton de Gruyter. Nielsen, Hans Frede. 1989. The Germanic Languages: Origins and Early Dialectal Interrelations. Tuscaloosa/London: University of Alabama Press. Osthoff, Hermann. 1876. Zur Geschichte des schwachen deutschen Adjecktivums. Forschung im Gebiet der indogermanischen nominalen Stammbildung, Vol. II. Jena: Pa¨ltz. Page, Richard B. 1998. Verner’s Law. Beitra¨ge zur Geschichte der deutschen Sprache und Literatur 122(2): 175–193. Petrova, Olga. 2004. The role of perceptual contrast in Verner’s law. In: Curzan and Emmons (eds.), 371–408. Prokosch, Eduard. 1927. The Old English weak preterits without medial vowel. Publications of the Modern Language Association 42(2): 331–338. Prokosch, Eduard. 1939. A Comparative Germanic Grammar. Philadelphia: Linguistic Society of America. Ringe, Don. 2006. A Linguistic History of English: From Proto-Indo-European to ProtoGermanic, Vol. I. Oxford: Oxford University Press. Stockwell, Robert P. and C. Westbrook Barritt. 1951. Some Old English Graphemic-Phonemic Correspondences – ae, ea and a. (Studies in Linguistics, Occasional Papers, 4.) Norman, OK: Battenburg Press. Thomason, Sarah, Gray and Terrence Kaufman. 1988. Language Contact, Creolization, and Genetic Linguistics. Berkeley: University of California Press. Vennemann, Theo. 2003. Bemerkung zum fru¨hgermanischen Wortschatz. In: Theo Vennemann (ed.), Europa Vasconica – Europa Semitica, 1–20. Berlin/New York: Mouton de Gruyter. White, David L. 2004. Why we should not believe in short diphthongs. In: Curzan and Emmons (eds.), 57–84. Wright, Joseph and Elizabeth Mary Wright. 1925. Old English Grammar. Oxford: Oxford University Press.
Jeannette K. Marsh, Waco (USA)
2 Periods: Old English
19
2 Periods: Old English 1 2 3 4 5 6
Preliminaries A chronological delimitation of Old English The external history and internal development of Old English Language-internal development: the decline of inflections Summary References
Abstract This chapter offers a survey of the main linguistic changes that took place during the Old English period – from the Anglo-Saxon migration around 450 CE to the beginning of the Norman rule of England. Considering that the major features and developments on all linguistic levels will be presented in Section III in more detail, the present article sketches the most salient and important linguistic features of Old English and otherwise focuses on political and cultural events of the period, which had an impact on the development of the English language. Predominantly, these are events that lead to the emergence of new contact situations – such as the Christianization (Latin), the Viking raids (Old Norse) and the emerging Norman influence on the English court in the 11th century (French) – and the most important waves of literary productivity – e.g. King Alfred’s educational program and the increasing book production following the Benedictine Reform.
1 Preliminaries The term “Old English” refers to those varieties of Germanic which were spoken in Great Britain from the Anglo-Saxon migration around 450 up until the end of the 11th century. While the geographical delimitation of the Old English language is unproblematic, the chronological limits are more difficult to determine and to some degree based on convention. Before I describe some of the main developments and characteristics of the Old English period, the most relevant approaches to and motivations for defining a chronological starting point and end point of Old English will be discussed briefly (Section 2). Taking into consideration that the main characteristics of the different domains of linguistic description are discussed in more detail in Section III, they will remain in the background in this article. The largest part of this chapter (Section 3) will focus on those aspects or developments of Old English that are related with or influenced by the non-linguistic history of its speakers. However, Section 4 will deal with languageinternal developments. It will be shown in this context that, while the choice of external dates for period boundaries may of course be associated with salient linguistic features during the development of a language, the relation between the internal and external factors is nevertheless mutual: once a choice of period boundaries has become conventional, the typological characterization of a language (in a given period) is dependent on this choice, which may be useful, but by no means necessary (cf. Curzan, Volume 2, Chapter 79). Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 19–32
20
I Periods
2 A chronological delimitation of Old English 2.1 The beginning of Old English In the case of Old English, it is easier to determine the beginning of the period than its end. Nevertheless, there are three historical events that can reasonably be interpreted as starting points in different respects. One is the settlement of Germanic tribes in England in the middle of the 5th century, the second is their Christianization around 600, and the third is the date of the earliest surviving written records around 700. With the arrival of Germanic settlers in England in the middle of the 5th century, their varieties of Germanic develop independently from the varieties of the cognate tribes that have remained on the Continent. Although the differences between the varieties of the settlers and those on the Continent cannot have been too great at the time of the migration, it is the settlers’ geographic and political independence as a consequence of the migration which constitutes the basis for the development of English as a variety distinct and independent from the continental varieties of the West Germanic speech community (cf. Section 3.1). Because close relations with the Continent persist for a relatively long time after the arrival of the Anglo-Saxons, it takes another one and a half centuries until their conversion to the Christian religion constitutes the first landmark of an independent AngloSaxon history. The Christianization and its impact on the Old English language will be discussed below in Section 3.2. At this point it may suffice to mention that, because the conversion is the first major change in the society and culture of the Anglo-Saxons that is not shared by the related tribes on the Continent, it is similarly significant for (the beginning of) an independent linguistic history of English as the settlement in Britain. Moreover, the immediate impact of the conversion on the language of the AngloSaxons is much more obvious than that of the migration: first, the Latin influence on English grows in intensity and, perhaps more crucially, enters new domains of social life; second, a new writing system, the Latin alphabet, is introduced, and third, a new medium of (linguistic) communication comes to be used – the book. Finally, one could approach the question of the starting point of Old English from a modern perspective. It is only indirect evidence that gives us a clue about the linguistic consequences of the two aforementioned events. Our direct evidence of any characteristic of (Old) English begins with the oldest surviving written sources containing Old English. Apart from onomastic material in Latin texts and short inscriptions, the earliest documents written in Old English date from the early 8th century. A distinction between a reconstructed “pre-Old English” before 700 and an attested “Old English” after 700, as drawn e.g. in Thomason and Kaufman (1988: 265), therefore does not seem implausible. While this criterion for determining the beginning of Old English is mainly based on a change in our modern access to the earliest stages of English (reconstruction vs. written evidence), some aspects of Anglo-Saxon history may in fact play a role here: as Hogg (1992b: 6) points out, it is feasible that the shift from a heptarchy of more or less equally influential Anglo-Saxon kingdoms to the cultural dominance of Northumbria in the time after Christianization may be connected with the fact that texts are produced not exclusively in Latin, but also in the vernacular. In other words, we may speculate (but no more than that) that the emergence of the earliest Anglo-Saxon
2 Periods: Old English cultural and political centre in Northumbria in the 8th century may lead the AngloSaxons to view themselves as one people rather than as different Germanic tribes, and, accordingly, to view their language as English (or, Anglo-Saxon) rather than as the Saxon, Anglian, Kentish, Jutish, etc. varieties of Germanic (but cf. an opposing point raised below in Section 3.1). Evidence for this change in attitude may be the composition of the Historia ecclesiastica gentis anglorum by Bede in 731, in which the gens anglorum of the title comprises all Germanic inhabitants of Britain and not only the Angles. Yet, as we can see from the earliest written evidence, the English language is not only sufficiently distinct from its closest cognates on the Continent around the year 700 with respect to its structure, its lexicon, and its phonology, it is at this point also considerably heterogeneous in itself – and continues to be so. I would therefore propose that, whatever happens to the language of the AngloSaxon settlers in Britain and for whatever reason it happens, any development after 450 should be taken as specifically English and before 450 should be taken as common (West) Germanic. That our knowledge of the underlying developments is necessarily based on a different method of access before and after around 700 is ultimately secondary to the relevant linguistic changes themselves and for any categorization of Old English.
2.2 The end of Old English The end point of Old English is marked by the Norman Conquest of 1066. The accession to the throne by William of Normandy in December of that year is considered a landmark in the history of England and, thus, of the community of speakers of English. It should, however, be questioned whether, or to what extent, the events of the year 1066 have only a symbolic value for the history of England rather than constituting an actual break. As will be discussed below (Section 3.6), in terms of the development of the contact situation between (Norman) French and English, the immediate relevance of the Norman Conquest is by far smaller than the prominence of this date in both the history books and in the handbooks on the history of English might suggest. The major linguistic changes that may be taken as relevant for a distinction between Old and Middle English take place in different linguistic domains – and, accordingly, at different times. The prototypical morphosyntactic features of Middle English – increasing syncretism of inflectional distinctions – begin to emerge as early as in the 10th century, whereas the typical Old English lexicon – largely Germanic with a moderate share of Latin borrowings but hardly any Romance elements – continues to exist up until the end of the 12th century and is attested even in Early Middle English texts such as The Owl and the Nightingale or Lagamon’s Brut; cf. Lutz (2002). In sum, if we define the Old English period as ranging from 450 to 1100 we mainly follow conventions. The distinction between Old and Middle English cannot be said to be motivated by sufficiently significant linguistic criteria and it is largely arbitrary to refer to the Norman Conquest in this context. Whether we determine 1066, 1100, or 1150 as the endpoint of Old English, it has become conventional to draw the line between Old and Middle English around this time. Strictly speaking, i.e. if we follow purely linguistic criteria, the transition from Old to Middle English expands over the period from the end of the 10th century to the end of the 12th century.
21
22
I Periods
3 The external history and internal development of Old English 3.1 The Anglo-Saxon migration The 7th-century historian Bede reports that in the year 449 Angles, Saxons, and Jutes, “de tribus Germaniae populis fortioribus” – ‘from the strongest tribes of the Germanic people’ (Bede, Historia Ecclesiastica 1.15; see Colgrave and Mynors [eds.] 1969: 50) – come to Britain and settle there. It has been mentioned above that the migration of the Anglo-Saxons means the beginning of a new speech community whose linguistic and non-linguistic history is now independent of the related tribes and their language(s). The main significance of the Anglo-Saxon migration therefore lies in the geographical reorganization of the Germanic speech community leading to the emergence of a new society and hence to a new, separate continuum of varieties. Although we do not know much about the social and political organization of these early Germanic settlers, some considerations are possible in this context. It is, on the one hand, reasonable to assume that the people involved in the migration speak varieties that originate in one and the same Continental Germanic dialect continuum. On the other hand, we do not know how close their varieties are or even to what extent there is mutual intelligibility among the settlers. But it is largely irrelevant how heterogeneous the language(s) of the Anglo-Saxons are at the time of their conquest. What we can reasonably assume is that, upon arrival in Britain, all the different groups involved in the settlement view themselves as speaking the same language in contradistinction to the Celtic inhabitants they encounter on the island. The identity of the settlers in their new homelands is necessarily based, among other things, on their common (albeit probably not quite uniform) linguistic background. The migration leads, moreover, to a new group identity of a subset of speakers of Germanic, who, irrespective of the heterogeneity of their own language(s), distinguish themselves both from the earlier inhabitants of Britain and from their relatives on the Continent. It is plausible to argue that it is this sociopsychological aspect which, more than anything else, constitutes the birth of the English language. Accepting this, we can assume, in turn, that it does not take too long after the settlement before the Anglo-Saxons view their version of Germanic as noticeably distinct from other Germanic varieties spoken by those who have stayed behind on the Continent. The migration itself does not immediately trigger any major change in the linguistic system. The earliest linguistic changes that English does not share with the cognate Germanic languages seem to be, at a first glance, independent of non-linguistic events. In the earliest period of Old English, a larger rearrangement of the phonological system takes place that affects mainly, but not exclusively, stressed vowels. (For a detailed description see Murray, Chapter 17; also cf. Campbell [1959: 50–112] and Hogg [1992a: 76–218] and for a shorter overview cf. Hogg [1992c: 100–119].) Two circumstances are employed in dating these sound changes. One is that other Germanic languages do not seem to have been affected by these changes. And secondly, the earliest written sources of Old English provide evidence that the relevant sound changes must have been completed by the date of their composition. Both these facts together suggest that all these sound changes take place in the time between 450 and 700.
2 Periods: Old English Although this dating is undisputed, the question nevertheless arises whether it is plausible to assume that, within the relatively narrow time frame between 450 and 700, the phonological system is considerably rearranged whereas it remains relatively stable for significantly more than 250 years both before and after this period. Any attempt to answer this question would have to remain speculative to some degree. But it is feasible to assume that the new social and political identity of the settlers leads to some cross-adaptations in the linguistic usage of the settlers. That dialectal differences among the settlers persist and, as we know, never cease to exist, does not exclude the possibility that some of the heterogeneity of these dialects is levelled out as a consequence of the formation of a new community. And since many of the obvious dialectal differences of a continuum are phonological, it is plausible that these adjustments take place predominantly on the phonological level. Bearing in mind the narrow time frame of the early Old English sound changes, it is therefore not implausible to assume that they have been enforced, if not triggered, by the formation of a new speech community as a consequence of the migration.
3.2 Christianization From around 600 onward, the Christian religion is spread quickly across the country both from the north, under the influence of the Irish Church, and from the south, by St. Augustine of Canterbury and his missionaries, sent to England by Pope Gregory I in 597. As indicated above, the Christianization of the Anglo-Saxons has immediate consequences on two domains of the English language – the lexicon and the writing system. Language contact with Latin is not a new phenomenon in those days. There has always been a moderate-to-intensive exchange of words between Latin and Germanic, ever since the two languages were neighbors on the Continent. Even after the migrationto England, the Anglo-Saxons adopt some Latin words, although the exact settings of these bilingual contacts are unclear. Plausible contexts are continuing relations with the other Germanic tribes on the continent, scattered speakers of Latin who have stayed in Britain after the Roman armies withdrew in 410, or speakers of Celtic who either use Latin as a lingua franca in communication with the Anglo-Saxons or whose language contains itself words borrowed from Latin which are then passed on to the Anglo-Saxons. But the words that are imported into the English language in the course of and after Christianization are of a considerably different kind than earlier, predominantly common-Germanic loans from Latin. While earlier Latin loans are words related to trading, to the military, or expressions for every-day concepts like household devices, the vocabulary imported with Christianization mostly denotes either concepts immediately related to the new religion and its institutions (e.g. abbod ‘abbot’, alter ‘altar’, munuc ‘monk’) or, generally, more abstract concepts. Another difference from earlier Latin loanwords is the medium through which they are introduced. Because many words come into English via books rather than via oral communication, the words are transferred from written registers of Medieval Latin rather than from Vulgar Latin, which is the source language of the pre-Christian borrowings. For the linguistic observer (rather than for the speaker of Old English), there is another notable difference. Earlier loanwords participate in most or all of the major sound changes that take place in the earliest period of Old English. That is, with respect to their phonology, the pre-Christian words from Latin have adapted to the
23
24
I Periods phonological system of Old English in their written attestations. By contrast, the more recent Latin words, introduced through Christianization, often retain a Latinate shape. The different phonological characteristics are best exemplified by Latin words that are borrowed both in the early and in the later period and that, accordingly appear in different forms in the Old English records. Thus, Lt. calix ‘chalice’ occurs in Old English in an older form, celc and a later form, calic. Likewise, OE leahtrice and OE lactuce are both borrowed from Lt. lactuca ‘lettuce’ at different times. (See Wieland, Chapter 23.) A far-reaching side-effect of Christianization is the introduction of the Latin alphabet in Anglo-Saxon England. Originally introduced as the medium on which the new religion is based, it soon comes to be used for the composition of other texts not immediately related to the Christian faith. Moreover, from about 700 onwards texts written in Latin script are composed in or translated into the English vernacular. Although never widespread in the early Middle Ages, through the introduction of the book and of Latin writing literacy enters new domains of social life which have not been reached by the use of the older writing system, the runic Futhorc. The first vernacular texts are mostly written in the north of England, in the monastic centers of Northumbria and Mercia. More than in the south, Christianity in the north is influenced by its Irish variety. The Irish use the Latin script themselves, although with a considerable number of stylistic alterations. Consequently, some of the Latin characters used in England in the early Middle Ages are of a slightly different shape than those used on the Continent. The ‘yogh’ or ‘insular G’, shaped 〈ᴣ〉, is used in Anglo-Saxon manuscripts to represent the voiced velar stop /ɡ/, the velar fricative /ɣ/ and the glide /j/. Only in the late Old English literature do we find instances of the ‘Carolingian G’, shaped 〈ɡ〉, then used to distinguish the stop from the fricative. The phonemes /æ/ and /æː/ are represented by an ‘ash’ 〈æ〉, a character shaped as a ligature of Latin 〈a〉 and 〈e〉. Other Latin letters used in Anglo-Saxon manuscripts differ considerably in shape from the variants we use today in printing. Three other characters are either not or indirectly taken from the Latin alphabet. Two of them certainly originate in the Futhorc, the set of Runic letters that is used by the Anglo-Saxons before the introduction of Latin writing: 〈ƿ〉 (‘wynn’) represents the bilabial glide /w/. Although the character 〈w〉 is usually employed in modern editions of Old English texts, Anglo-Saxon manuscripts use 〈ƿ〉 almost throughout, and only occasionally 〈u〉 or 〈uu〉 for /w/. When Old English is represented by the Latin alphabet, ‘thorn’ 〈þ〉 and ‘eth’ 〈ð〉 are used indistinguishably as allographs for both the voiced and the unvoiced dental fricatives, /θ/ and /ð/. While the former is also adopted from the Futhorc, the origin of the latter is uncertain (cf. Hogg 1992c: 75).
3.3 The Vikings in England The raid on the Lindisfarne monastery in 793 is the first known instance of a series of increasingly intense attacks on England by Vikings. While initially local plundering rather than attempts at gaining political influence motivates the raids, their quality and purpose change, and by the middle of the 9th century, a large area of England comprising almost all parts of Northumbria and Mercia is under Danish control. When King Alfred of Wessex comes to the throne in 871, his West Saxon kingdom is the only autonomous area of what once used to be the Anglo-Saxon heptarchy. Alfred succeeds in protecting his own territory from Danish rule and also in re-conquering the western
2 Periods: Old English parts of Mercia. Moreover, as a consequence of his military success he manages to negotiate a truce resulting in the Treaty of Wedmore in 878. In this treaty Alfred and the Danish leader Guthrum agree on a borderline between an area of Danish legislation (the Danelaw) and an independent Wessex. The linguistic consequences of the Viking rule in England are difficult to measure. Because of the division of England into a Danish and an English political sphere, the situation in the 10th century is as follows: in the south, a relatively stable and peaceful political situation allows Alfred to instigate an educational reform. He supports the import of new books from the continent and the production of new books in England, and he also initiates the translation of Latin books into English and the production of books written in the vernacular. Thus, for the remaining two centuries of the AngloSaxon period, the vast majority of sources comes from the area of the West Saxon kingdom. In the north, by contrast, Viking influence on English is naturally much stronger. Therefore, judging from surviving Old English texts alone, the evidence of language contact between Old Norse and Old English in England is quite small – about 150 words of Scandinavian origin are attested in the Old English sources. Judging however from Middle English evidence, we may assume that there must have been a very intensive contact situation, at least in the area of the Danelaw (cf. Kastovsky 1992: 320). This can be seen from the type of loanwords and their features, rather than from the mere number of words of Scandinavian origin in the English of today. In contrast to the Latin loanwords introduced through the influence of Christianization, Scandinavian loanwords in English are less technical and much more part of the basic vocabulary of English. Words like skirt, egg, or sky, for example, denote everyday concepts rather than more abstract and learned concepts as represented by the Latin borrowings. Much more efficiently than Latin words, Old Norse loanwords seem to have been integrated into the basic vocabulary with a high token frequency. An example is Old Norse tacan ‘take’ which replaces the highly frequent Old English niman and which is one of the few borrowed verbs that is (even today) inflected as a strong verb. Similarly, the verb get is, if not borrowed from Old Norse, at least influenced in its phonological shape, as the Old English equivalent gietan was pronounced with an initial glide, i.e. /'jetɑn/, and would have resulted in PDE *yet rather than get. Moreover, more than the French influence during the Middle English period, Old Norse at some points enters grammatical structures of English, as we have traces of Old Norse in the pronominal paradigms (e.g. they, them, their) and in the inflectional paradigms (e.g. -s 3P SG for southern -[e]þ). It should also be noted that Old Norse and Old English are quite close cognates. It is impossible to say whether or to what extent the two languages are mutually intelligible in the 9th and 10th centuries, but many words seem close enough. Townend argues that there is no full comprehensibility between Old Norse and Old English, but that there is what he calls “adequate” or “pragmatic intelligibility” (Townend 2002: 181–183), i.e. a degree of comprehension that allows for basic conversation, but that does not cover the full morphosyntactic complexities of the two systems. Whether or not two etymologically equivalent lexemes can be confused, it is more crucial that, in some cases, their phonological differences cannot be distinguished by the spelling conventions of Old English. What matters therefore is to what degree, given the close relatedness of the two languages, the spelling conventions of Old English allow us to identify traces of
25
26
I Periods linguistic transfer. In other words: how can an Old Norse loanword be distinguished from its Old English cognate in a nearly contemporary source? For instance, the fricative /ʃ/ is represented in Old English manuscripts by the digraph 〈sc〉, but this is at the same time the combination of letters which serves best to represent the Old Norse cluster /sk-/. As a consequence cognate pairs like Old Norse /'skyrte/ (PDE skirt) and Old English /'ʃyrte/ (PDE shirt) are impossible to distinguish by the Old English spelling: both forms would appear as 〈scyrte〉, and it is therefore impossible to say when the Scandinavian loanword skirt entered the English language.
3.4 Alfred’s educational reform King Alfred’s contribution to the history of the English language is twofold. One aspect is that, as a result of his military success, the Danish conquest of England comes to a halt (Section 3.3). The claim that Old Norse would have become the major language of England if the Danish troops had also occupied the south of the country must necessarily remain speculative, but it can at least be assumed that Old English would have been a threatened language and that, had Alfred not defended a stable (in political and military terms) English-speaking area, English would have lost its role as a predominant language in England. There is another achievement of King Alfred which may not influence the development of English with the same intensity as the contact with Old Norse does, but which influences considerably our knowledge of the English language in the early Middle Ages. Alfred’s educational reform is the impetus for a considerable increase in the production (and import) of books in general, and in the production of written literature in Old English in particular. Alfred himself initiates the translation of a number of important and influential Latin texts: Gregory the Great’s Cura Pastoralis and Dialogi, Augustine of Hippo’s Soliloquia, Boethius’s De consolatione philosophiae, Paulus Orosius’s Historiae adversus paganos and, perhaps most important of all, Bede’s Historia ecclesiastica. The degree of Alfred’s personal participation in the translation process varies (and remains disputed), but it can be said that all these translations result from his education policy. Moreover, a number of vernacular texts are composed in the same context, again with varying degrees of Alfred’s personal involvement, e.g. the Anglo-Saxon Chronicle, which is continued in several versions up until the 12th century, and a Martyrology. As a result, up until the end of the Anglo-Saxon period the vast majority of texts of various genres is produced in the royal court or in the monastic centers of the West Saxon kingdom, most of all in the capital Winchester. For the first time in the history of Old English a large corpus of long prose texts is produced, and for the first time there is evidence of a more standardized written language mainly based on features of the West Saxon dialect, but not without traces from the variety attributed to Mercia (cf. Section 3.5). Up to the end of the Old English period, a large number of Old English documents originate in West Saxon or are heavily infiltrated by features of the West Saxon dialect. The evidence from the Old English sources for our knowledge of the history of the English (spoken) language is therefore clearly misleading, because the varieties spoken in the Midlands (i.e. what in Anglo-Saxon times was the Mercian variety) contribute much more to the development of Present Day (Standard) English than those of West Saxon do.
2 Periods: Old English
3.5 The Cluniac reform Half a century after Alfred’s reign, the flourishing book production receives a further impulse. This is triggered by a movement that affects the ecclesiastical history of England rather than, as in the case of Alfred’s contribution, the political history, although the two are tightly connected. A monastic reform movement aiming for a stricter and more ascetic interpretation of the Benedictine Rule initiated in the monasteries of Cluny and Fleury (later Saint-Benoıˆt-sur-Loire) in France spreads to England in the middle of the 10th century. One of the central figures in the Cluniac reform movement in England is Æthelwold, Bishop of Winchester 963–984. Educated at the court of King Æthelstan (reigned 921–939), one of Alfred’s brothers and successors on the West Saxon throne, Æthelwold continues what Alfred has begun. Although Alfred and Æthelwold carry out two different reform projects, both instigate a revival of literary and intellectual productivity after a period of a constant Viking threat has, at least in this region of England, come to an end. While Æthelwold’s commitment is religious rather than political, he can be sure of royal support for his work. The school founded by Æthelwold at the New Minster in Winchester produces not only books in high numbers. The texts that emerge from the scriptorium at Winchester also have a remarkable stability in orthography and also, as far as we can judge, in their vocabulary. From this, it has been deduced that the language of many documents composed or copied in Winchester and in related scriptoria represents the first attempt at a standardization of the English language. It is, on the one hand, obvious that West Saxon texts from the end of the 10th century onwards often show a remarkable uniformity in the choice and orthographic form of words. But it should also be taken into consideration that we are dealing with a set of texts covering a limited range of scholarly fields. It would be problematic to deduce the existence of a genuine standard language from the relative homogeneity of the Winchester texts alone. Indeed, the very fact that the documents representing “Standard Old English” all derive from a tight network of authors and instigators in a predominantly monastic context – all in all a rather small, albeit influential, group of people – speaks against rather than in favor of the wider use of their linguistic features outside these circles. It is therefore justified to speak of orthographic conventions characteristic of the Winchester school, perhaps of a West Saxon Schriftsprache, but it is difficult, if not impossible to make judgments about the scope and influence of the Winchester conventions. The impression that the “Winchester standard” spread widely is not only owed to the political situation described in the Section 3.4, i.e. to the fact that other dialect areas were to a larger or smaller extent excluded from the production of books – at least judging from the material that has survived until today – it is also due to the fact that the most productive single author of the Anglo-Saxon period, Ælfric of Eynsham (c.955–c.1010), is a disciple of Æthelwold. Therefore, it is perhaps more appropriate to say that a large share of West Saxon texts of the later Old English period (and thus of the corpus of Old English texts in general) has some roots in the Winchester tradition rather than that the Winchester tradition contributes to the standardization of Old English. A few general caveats should be expressed in this context. First, not many of the texts that have survived until today can be said to be pure representatives of one specific variety. Perhaps such a statement would apply most to prototypical representatives of the variety of 10th century Northumbria, such as the Lindisfarne Gospels and the
27
28
I Periods Durham Ritual. But pure West Saxon texts are rare, for several reasons, such as the fact that many of the West Saxon texts, including Winchester documents, are copies of older, lost versions and attest to an earlier layer of Mercian linguistic features. Second, West Saxon cannot have been a dialect as uniform as the descriptions of the Old English dialects often suggest. Kastovsky (1992: 346) points out that the traditional distinction between “Early West Saxon” for the language prototypically represented by the Alfredian translations and “Late West Saxon”, prototypically represented by Ælfric’s texts, is misleading, as it suggests a mere diachronic distinction between two varieties within only a little more than a hundred years. Rather, it should be assumed that the differences between the two groups of documents are diatopic at least to the same extent as they are diachronic. Third, as indicated above, the idea of an Old English standard language presupposes not only a process of deliberate regulation; it also requires a broader distribution of a standard language in larger parts of the population. As to the last point it should be noted that perhaps the term “standard”, introduced in this context by Gneuss (1972), is the main problem in this context. If a standard is understood as an institutionalized variety that, among other things, serves as a means of communication bridging several local and social differences in the usage of a language, the hypothesis of a Late West Saxon standard involves two problems. First, it is not falsifiable, because we have no clues as to how widely a deliberately regulated variety may have made its way outside the scriptoria. And second, the idea is implausible because it is not clear how a variety attested in a number of specialized scholarly texts should have spread into other areas of society given that literacy was limited to a rather small elite. What is plausible, though, and for this we do have evidence, is that there is an influential intellectual elite which has an enormous impact on the literary productivity in late AngloSaxon England, and who seem to have used the language of their works in a deliberate and comparatively uniform way. (See further Kornexl, Chapter 24.).
3.6 Old English and Old French In addition to the impetus for the production of literature, the import of the Cluniac reform into England must also be seen in a different context of the history of English. It has briefly been discussed earlier (Section 2.2), that French influence on English begins gradually and not abruptly with the Norman Conquest. The monastic reform is in fact the first instance of contact between speakers of French and English. That the Gallo-Romance vernacular is perceived as sufficiently distinct from Latin can be deduced from the explicitly trilingual character of the Oaths of Strasbourg of 842. Although the Oaths can hardly be employed as a data source for the Old French language, it nevertheless attests to the fact that the French vernacular is considered an idiom independent from any variety of Medieval Latin. That this applies not only to the perspective of the speakers of early Old French but also to that of the Anglo-Saxons around the turn of the millennium is confirmed by an English source composed in 1011 by Byrhtferth of Ramsey, i.e. his Manual or Enchiridion, in which Byrhtferth makes a remark on the correct use of Latin versus French (Byrhtferth, Enchiridion 2.1. 449–454; see Baker and Lapidge [eds.] 1995: 88–90). In spite of the fact that French and Latin were without doubt two distinct idioms in the late 10th century, it is nevertheless impossible in many cases to distinguish clearly whether Romance material in English documents is of Latin or of French origin. For
2 Periods: Old English this reason alone there are hardly any clear traces of contact between French and English. Yet, traces do exist: particularly in 11th-century (but pre-Conquest) glossaries, we do occasionally find French words among the Old English interpretations. Most of them, provided they are unambiguously of French rather than of Latin origin, are attested only once, so that we cannot assume that they have ever been part of the English lexicon. Only two such words are clearly French and are attested more than once: capun ‘capon’ in the Antwerpen and Brussels glossaries and iugelere ‘magician’ occurring several times in different glossaries to Aldhelm’s De laude virginitatis and, notably, once with a different spelling in an anonymous homiletic text, i.e. not as a gloss but in a prose text (cf. von Mengden 1999). The period of the monastic reform is certainly the earliest date from which contact between speakers of the two languages is attested. It is plausible, therefore, to assume at least a slight degree of lexical transfer. The contact situation continues in the early 11th century when the relations of the Crown with the Duke of Normandy intensify, at the latest under the reign of Æthelred II (reigned 978–1016), who married Emma, the sister of Richard II, Duke of Normandy. That is, in both royal and ecclesiastical circles, there are tight connections between French and English speaking people around the year 1000. It is difficult to determine how far-reaching and how widespread contacts between these two groups are in England in the first half of the 11th century. But there is some evidence of Norman influence in pre-Conquest England. The occupation of official positions by native speakers of (Norman) French began, albeit on a small scale, as early as the 1040s, with Robert of Jumie`ges being appointed Bishop of London in 1044 (and promoted to Archbishop of Canterbury in 1051). And both of the two separate entries for the year 1052 in the Worcester Chronicle refer to a Norman castle in England (ChronD 1052.1 15 and 1052.2 2; see Cubbin [ed.] 1996: 70, l. 19 and 71, l. 6). These passages clearly imply that the Normans must have been numerous and powerful enough to erect their own military fortifications on English soil – some fifteen years before the Conquest. But this passage in the Chronicle at the same time attests to ongoing language contact: the expression used to refer to the fortification is in fact the first instance of the English word castle, in its Old English form castel. Its meaning and its grammatical gender (MASC) reveal that it must be a borrowing from Norman French and that it cannot be identical with the homonymous Latin loanword castel (NEUT) ‘town, village’. One should, of course, not overstate the linguistic transfer between French and English before the Conquest. But these aspects may suffice to support the point made above in Section 2.2, i.e. that the Norman Conquest as such did not have any immediate consequences for the English language. The events of the year 1066 seem to have been the consequence of a series of steps by the Norman nobility to gain political influence in England – a development always accompanied by support from an influential pro-Norman party in the Anglo-Saxon aristocracy. It is therefore feasible to assume that the intensity of French influence, although traceable, is not considerably greater in the years immediately following the Norman Conquest than it is before. From this perspective, the Norman Conquest stabilizes, but by no means ignites or reinforces, the growing intensity of Anglo-Norman relations. As such, William’s victory at Hastings may be seen as one of several important events that pave the way for the enormous influence that French exerts on English in the 13th and 14th centuries. The beginnings
29
30
I Periods of this development are clearly part of the history of Old English rather than of Middle English.
4 Language-internal development: the decline of inflections The previous section focused on those developments that were either triggered by language-external events or, at least, should be seen and explained in the context of the history of the speech community. A selection of instances of internal change taking place during the Old English period will be discussed briefly in the following. Old English is often described as an inflecting language. This label follows a particular classification according to morphosyntactic types of languages, which can be observed cross-linguistically. It is particularly prominent in the descriptions of Old English because it is motivated by the contrast with the analytic character of Present Day English. In this context, the classification goes back to Henry Sweet (1874: 160) who took the “full inflections” of Old English as the main defining criterion for his periodization of the English language (cf. the discussion in Curzan, Volume 2, Chapter 79). According to this approach, Middle English is the period with a “limited” set of inflectional categories. The decline of a complex system of case inflections begins, however, long before the period that we have defined as Old English. During Old English times, the dual number is retained only in some pronominal forms and disappears almost completely by the end of the Old English period. Case syncretism has been a continuous process at all times. While for Proto-Indo-European eight cases are reconstructed, all daughter languages of IndoEuropean have less than eight cases even in their earliest attested stages. At the beginning of the literary period of Old English, the merger of instrumental and dative has almost been completed, with distinctions retained only in some pronominal forms and in a few adjectives. But to say that during the Old English period a system of formerly five cases reduces to four cases, would again be a simplification, because in many forms of masculine and feminine nouns (neuter nouns never encode the distinction), nominative and accusative are not any longer distinguishable in 11th-century sources. Thus, in spite of the categorization as the “period of full inflections”, during the Old English period as many distinctively encoded case values get lost as in Middle English. The causes of this particular stage in the reduction of the case system are predominantly phonological. The fixed, initial stress characteristic of the Germanic languages generally entails unstressed final syllables. The result is that first front vowels begin to merge in final syllables in the earlier stages of Old English, a process naturally affecting many inflectional endings. Only in the 11th century does the phonetic reduction also affect back vowels. Thus the dative plural suffix -um is comparatively stable and even in its reduced forms /-on/ or /-ən/, it is still distinguishable from the other case/gender suffixes in most noun classes because of the nasal; cf. Hogg (1992a: 3n. 2). If we consider that the loss of case distinctions necessitates the (ultimately Middle English) replacement of the predominant V2 word order in Old English by a rigid SVO order, we can observe a long term development of cross-influences of various linguistic domains: intonation (Germanic initial stress) → phonology → inflectional morphology → syntax. Of these, Old English particularly observes the phonological changes, the syncretism of some of the inflectional markers and also the reduction (or loss) of some inflectional categories.
2 Periods: Old English The decline of morphological values during the Old English period is more substantial in the nominal paradigms than in the verbal forms. While verbal endings are affected by the phonological reduction too, the main difference between the nominal and the verbal system is that the reduction of verb endings does not result in a syncretism of inflectional values to the same extent to which it does in the nominal paradigms. Throughout the Old English period three person values are distinguished in the singular, but not in the plural. There is a general tense distinction between past and present. Finally an indicative and a subjunctive mood are distinguished morphologically. The most salient feature of the Old English verbal system is shared with all other Germanic languages: the distinction between strong and weak verbs. While weak verbs mark their inflectional values by suffixes, strong verbs use a combination of suffixes and systematic vowel alternations. In this context, the typical descriptions are again more idealized for Old English than they are for later stages of English. The relatively clear set of seven classes of strong verbs that we often find in handbook descriptions is, naturally, full of idiosyncrasies. Moreover, the traditional class distinctions are based on a set of criteria which are not completely consistent (cf. the more detailed discussion in von Mengden, Chapter 18, Morphology, Section 2.4). Again, the impression that the Old English system of strong verbs appears to be more regular and systematic than the Middle English system is certainly not wrong. Yet, it should be noted that handbook descriptions of the verbal system of Middle English tend to include variation among and within the paradigms of the verb classes, whereas equivalent descriptions of Old English focus more on their regularity. Therefore, although we can generally assign the label ‘inflectional language’ to Old English – irrespective of its wide diachronic and diatopic variation – it should at the same time be borne in mind that the decline of inflectional categories has been a continuous process since long before the Anglo-Saxon migration. If we refer to Old English as the period of “full inflections” (Sweet 1874: 160), the attribute “full” wrongly implies that the Old English system has reached the highest possible degree of morphological complexity – both from the point of view of the history of English and from a cross-linguistic perspective. Rather, the factors which justify assigning a new label – “English” – to a language whose history does not really have a beginning, and on the basis of which we have defined a starting point of Old English in Section 2.1, are of a sociolinguistic nature. But neither the beginning nor the end of Old English coincide with any salient changes in the inflectional system (cf. the discussion in Curzan, Volume 2, Chapter 79).
5 Summary Bearing in mind the various problems involved in both periodization and categorization discussed above, we may say that Old English is, particularly in contrast to later stages of English, a typically Germanic language in many respects. The share of inherited Germanic words in the vocabulary is much greater than it is today, and even the moderate share of Latin loanwords is a feature that applies to all Germanic languages in the early Middle Ages. Its syntactic (V2) and morphological (inflectional) features are similarly characteristic of Germanic. Finally, in spite of a major rearrangement of stressed vowels at an early stage of the Old English period, the phoneme inventory, too, is basically the same as that of the other early Germanic languages.
31
32
I Periods
6 References Baker, Peter S. and Michael Lapidge (eds.). 1995. Byrhtferth’s Enchiridion. Oxford: Oxford University Press. Campbell, A. 1959. Old English Grammar. Oxford: Clarendon. Colgrave, Bertram and R. A. B. Mynors (eds.). 1969. Bede’s Ecclesiastical History of the English People. Oxford: Clarendon. Cubbin, G. P. (ed.). 1996. The Anglo-Saxon Chronicle: A Collaborative Edition. Vol. 6: MS D. Cambridge: Brewer. Gneuss, Helmut. 1972. The origin of Standard Old English and Æthelwold’s school at Winchester. Anglo-Saxon England 1: 63–83. Hogg, Richard M. 1992a. A Grammar of Old English. Vol. 1: Phonology. Oxford/Cambridge, MA: Blackwell. Hogg, Richard M. 1992b. Introduction. In: Hogg (ed.), 1–25. Hogg, Richard M. 1992c. Phonology and Morphology. In: Hogg (ed.), 67–167. Hogg, Richard M. (ed.). 1992. The Cambridge History of the English Language. Vol. 1: The Beginnings to 1066, 1–25. Cambridge: Cambridge University Press. Kastovsky, Dieter. 1992. Semantics and vocabulary. In: Hogg (ed.), 290–408. Lutz, Angelika. 2002. When did English begin? In: Teresa Fanego, Bele´n Me´ndez-Naya, and Elena Seoane (eds.), Sounds, Words, Texts and Change: Selected Papers from 11 ICEHL, Santiago di Compostela, 7–11 September 2000, 145–171. Amsterdam/Philadelphia: John Benjamins. Sweet, Henry. 1874. A History of English Sounds from the Earliest Period. Oxford: Oxford University Press. Thomason, Sarah Grey and Terrence Kaufman. 1988. Language Contact, Creolization, and Genetic Linguistics. Berkeley: University of California Press. Townend, Matthew. 2002. Language and History in Viking Age England: Linguistic Relations between Speakers of Old Norse and Old English. Turnhout: Brepols. von Mengden, Ferdinand. 1999. Franzo¨sische Lehnwo¨rter im Altenglischen. In: Wolfgang Schindler and Ju¨rgen Untermann (eds.), Grippe, Kamm und Eulenspiegel: Festschrift fu¨r Elmar Seebold zum 65. Geburtstag, 277–294. Berlin/New York: Mouton de Gruyter.
Ferdinand von Mengden, Berlin (Germany)
3 Periods: Middle English 1 2 3 4 5 6 7
What is Middle English? Issues of evidence Writing and speech Grammar Lexicon Summary References
Abstract Middle English is the period in the history of English when variation is most thoroughly recorded in the spoken mode, and the body of surviving material is very large. In this chapter, the extralinguistic reasons for the survival of variation in Middle English are Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 32–48
3 Periods: Middle English given, focusing on the relationship between linguistic form and socio-cultural function. Middle English has come down to us directly through manuscripts, and indirectly through reconstruction and through the study of residualisms in Present-day English; the reliability of the evidence for Middle English is assessed. Each level of language in turn is then discussed: writing- and speech-systems (= transmission), grammar and lexicon. New resources (especially online) for the study of Middle English are flagged, offering exciting possibilities for future research. The aim of the article is to offer both a characterization of Middle English itself and a sketch of the resources available for its study.
1 What is Middle English? Middle English is the form of English spoken and written roughly between 1100 and 1500 CE. Its beginning roughly corresponds to the Norman Conquest of 1066, while its end roughly corresponds to the first book printed in English (1475) and the arrival of printing in England (1476): these two historical events, though of very different kinds, have implications for the status of the language during the Middle English period. Up until the Norman Conquest, English (i.e. Old English) had a distinct status as a language of record and for literary expression; it developed a written standardized form, classical Late West Saxon, which was used by scribes outside its area of origin, i.e. Wessex in south-west England (see Kornexl, Chapter 24). After the Norman Conquest, Middle English was displaced in prestige for literary and documentary manuscripts by Latin, and as a language of literary culture (and later record) by varieties of French: by Norman French (which in England developed into Anglo-Norman) up to the middle of the 13th century, and by Central French after that. English remained the speech of the majority of the population and continued to be widely written; but the written mode had a restricted function. If writers wished to communicate beyond their area of origin, or wished to leave a record for subsequent generations, they tended to use other languages than English. These restricted functions for English had linguistic implications, as we shall see. Towards the end of the Middle English period, English became more elaborated in social function, i.e. used more commonly as a language of record and literary function. The process was gradual: Latin remained common as a language of record and of “serious” literature well into the Early Modern period, and there are still remnants of Anglo-Norman in certain set phrases in the British parliament, e.g. La Reyne le veult ‘the Queen wishes it’ for public bills on receiving royal assent (even though parliamentarians began using English from the middle of the 14th century). But the efflorescence of English literature which we associate with (say) the Elizabethan and Jacobean theater could only have come about because it became possible to use English in a more elaborated way. Reasons for this change include the Reformation, which gave the vernacular a new role in religious expression, and the rise in power of the “middling” classes, notably the bourgeoisie, for whom English was the usual mode of expression (see Schendl, Chapter 32). Defining Middle English in linguistic terms is in some ways straightforward, with reference to every level of language traditionally identified: in lexicon, in grammar, in sound-system, in writing-system (see Section IV). Middle English differs from Old English in that it manifests large-scale borrowing of vocabulary from French, and in that Norse vocabulary, which had been mostly a feature of the spoken language in
33
34
I Periods Old English, appears in the written record. In grammar, Middle English is often referred to as the period of “reduced” inflexion, whereby the relationship between words and phrases is expressed to a greater extent than in Old English through element order, the use of prepositions, and through markedly distinctive pronouns; however, there is a wider range of inflexions than in Present-day English, notably with regard to the verb but also (in many varieties) in the adjective. A phonological distinction between voiced and voiceless fricatives, not made in Old English, appears in Middle English; and there are changes in the vowel-system, most notably in the loss of quantity as a feature distinguishing “short” and “long” vowels and in the configuration of the system of diphthongs. These developments are in many cases reflected in the writing system of Middle English, which is highly variable, in that spelling usages seem to have varied from parish to parish; variation in every level of language is reflected in writing. The variety of written Middle English clearly relates to its social function; if English had a local rather than a national function, it makes sense for written forms of Middle English to reflect the spoken mode quite closely, to help readers who would encounter it primarily at the first stage of literacy. Middle English similarly differs from Modern English. In vocabulary, the transition from Middle to Modern English saw the transfer of large numbers of items not only from French and Latin but also the beginnings of a flood of words from non-European languages, the result of trade and the beginnings of imperial expansion. During the transition from Middle English to the Early Modern English period, usually dated 1500–1700, the present-day pattern of inflexion emerged, while the sound-system underwent a radical change, most notably with regard to the long vowels of late Middle English: the Great Vowel Shift (see Krug, Chapter 48). Finally, variety in writing systems was gradually suppressed as the Middle English period segued into Early Modern English, first of all in writings designed for public consumption and eventually in private letters as well. Again, this suppression of variety related to the changing function of English; it made sense for a written language used for the purposes of record or national communication to develop a form which was less prone to variation. But such a straightforward characterization of Middle English, though useful as an initial outline, is a massive over-simplification of a highly complex period in the evolution of the English language, omitting the extralinguistic contexts which triggered these changes and the dynamic interaction of levels of language which took place at different speeds, and for different reasons, in different parts of the country. Old English did not become Middle English at midnight in the year 1100; Middle English did not become Modern English at midnight in the year 1500. Moreover, it is in many ways more accurate to consider the history of Middle English – as arguably of all English – as the history of Englishes, the history of (in Roger Lass’s useful phrase) “a population of variants moving through time” (Lass 2006: 45). The over-simple characterization offered at the beginning of this section also omits something else crucial: the fact that the primary evidence for Middle English is, though extensive, nevertheless limited. We know a great deal about Middle English – indeed, we are learning more about this state (these states?) of the language all the time – but there is also a great deal which we clearly do not know, and which it is likely we never will. In sum, the characterization just offered is a useful listing of prototypical features but not a history, i.e. a narrative which offers at least a partial explanation for – or, in April McMahon’s phrase, a “relief from puzzlement” about – the phenomena described
3 Periods: Middle English (McMahon 1994: 45). This chapter is designed to engage with current research in this field.
2 Issues of evidence The primary evidence for Middle English is much restricted in comparison with the massive resources available for students of Modern English. There are, of course, no soundrecordings of Middle English speech, which have been available to students only since the end of the 19th century; and there are no contemporary attempts at structured linguistic descriptions, as are available for scholars working in the Early Modern period. Students of Middle English interested in engaging with contemporary linguistic scholarship on the English language have to make do with the somewhat vague and general comments about linguistic differences made by literary authors, mocking the usages of folk from parts of the country other than their own, e.g. Chaucer’s representation of Northern speech in the Reeve’s Tale, or the humorous representation of Southern speech in the Wakefield Second Shepherd’s Play. There is, unfortunately, nothing in Middle English equivalent to the Old Icelandic First Grammatical Treatise, the sophisticated treatise on pronunciation of the vernacular produced in Iceland in the 13th century. Evidence for Middle English consists of material remains, most notably manuscripts written by scribes on parchment and latterly on paper, and (at the very end of the Middle English period) early printed books. Some texts – more than have until recently been recognized – were written on the walls of churches, especially towards the end of the period. Thousands of these objects have survived, even though it is likely that very many have been lost. Such texts were produced not only in the scriptoria of monasteries or cathedrals but also in private households, country vicarages and – increasingly so towards the end of the period – in workshops in towns, notably London. Many are now stored in the great libraries of the world: the British Library in London; the college libraries of Oxford and Cambridge; the Bodleian Library in Oxford and Cambridge University Library; the great private libraries of the United States, such as the Huntington and the Pierpont Morgan, and the libraries of universities such as Princeton and Yale; the libraries of Trinity College Dublin and the University of Durham; the John Rylands and Chetham Libraries in Manchester; the Bibliothe`que Nationale in Paris; Glasgow’s Hunterian Collection and Edinburgh’s National Library of Scotland. But there are also many manuscripts in less obvious places: in (e.g.) the private Takamiya collection in Tokyo, or in the muniment rooms of local councils, or, even today, of “great houses” such as Alnwick or Petworth. These manuscripts and early printed books contain texts ranging from major literary works (poems such as Chaucer’s Canterbury Tales or Langland’s Piers Plowman, or works of religious instruction such as sermons and saints’ lives) to more pragmatic and/or ephemeral material (e.g. works of medical reference, wills, letters). Such texts, of course, survive not through conscious selection of representative material, but by chance, often because of the idiosyncratic interests of the 16th- or 17th-century antiquaries who collected them, such as Sir Robert Cotton (whose manuscripts are now in the British Library) or Franciscus Junius (whose collection is now in the Bodleian). Whereas modern sociolinguists or dialectologists can select their informants on the basis of age or social class or precise locality, students of Middle English have to make do with what the vagaries of time have bequeathed to them.
35
36
I Periods Thus, the first task for any student of Middle English is always to assess the status of the primary evidence presented: the philological enterprise of textual analysis must precede engagement with larger questions of linguistic enquiry. Traditionally the philological study of Old and Middle English has been expressed through the editing of texts: “critical” and “diplomatic”. The critical edition has dominated textual scholarship since at least the middle of the 19th century, notably in the editing of Biblical and classical texts, and is generally referred to as “textual criticism”. The goal of textual criticism was the creation of the “critical edition”, whereby the “threads of transmission” were “follow[ed] back […] to restore the texts as closely as possible to the form which they originally had” (Reynolds and Wilson 1974: 186). The principles of textual criticism, which emerged first in the editing of Biblical and classical texts, were transferred to the editing of vernaculars, and some of the most sophisticated textual criticism has been carried out by editors of Middle English texts. Thus, the editors of the great Athlone edition of Piers Plowman, one of the principal (if controversial) achievements of 20th-century Middle English editorial scholarship, refer to the outcome of their enterprise as follows: “a theoretical structure, a complex hypothesis designed to account for a body of phenomena in the light of knowledge about the circumstances which generate them” (Kane and Donaldson 1974: 212). The focus of the critical edition, therefore, is the author’s intention; it is no coincidence that textual criticism flourished as the authorial voice became reified in Romantic and Victorian cultures, and critical editions, published in prestigious series such as those issued by the Early English and Scottish Text Societies (EETS, STS), continue to dominate medieval English literary studies. Textual criticism has developed a battery of techniques whereby extant materials are compared (“collation”), and whereby putative errors – more properly, non-original readings – are identified and used to establish relationships between texts. Once these relationships have been expressed through an appropriate modeling process, traditionally the “stemma”, the archetypal text is deemed to be established and can then be corrected in the light of the critical editors’ judgement, based on their knowledge of the linguistic and cultural contexts of the period in question. There is, of course, a clear parallel with the tree-diagrams of the 19th-century historical linguists, and it is no surprise that many such linguists combined the editorial and linguistic enterprise (thus the traditional meaning of the term “philology” in Anglophone countries). With the rise of new approaches to literary criticism from the 1930s onwards, however, there was a decoupling of the literary and linguistic traditions, and most recent EETS and STS publications are critical editions (albeit “conservative” ones), designed primarily for literary scholars and focused on the authorial intention for the work in question. Such editions are potentially problematic for the linguistic student of Middle English, since they are avowedly an abstraction from the surviving data, which is in almost all cases mediated through scribes. It is therefore often necessary for linguists to return to the original materials, but medieval manuscripts are of course restricted in terms of access; their fragility means that they cannot be easily transported from place to place, and there are also conservation issues to which librarians are rightly sensitive. Photographic images are thus commonly used, traditionally on microfilm or in published facsimiles but increasingly on the internet (for a useful resource, see the relevant section of the Middle English Compendium [McSparran (ed.) 2006]). However, photographic reproductions present
3 Periods: Middle English linguists with further problems. Publication of images, however accurate, does not offer any interpretation of the facts presented. For such an interpretation, “diplomatic” editing, demanding high-level hermeneutic skills (paleographical, codicological, linguistic) are needed: “What is necessary is that a single editor should spend the time necessary to solve the problems of the manuscript, even those that are themselves trivial and unimportant, and should find a means of presenting his [sic] results so that others may benefit from his [sic] pains; the job should be done thoroughly once, not superficially by each individual user of a facsimile” (Dobson 1972: xii–xiii). Diplomatic editions, especially when presented in machine-readable form, also allow for “text-mining”, whereby texts are searched for forms of interest and the results analysed using various statistical tools. Diplomatic editions are not often favored by literary scholars, and it is noticeable that EETS, for instance, has not produced diplomatic editions to follow on from the Ancrene Riwle series. Yet there are modern developments in what might be termed “computational philology” which are beginning to recuperate the diplomatic edition in Middle English studies, at least for those whose scholarly orientation is primarily linguistic. Computational philology applied to English historical linguistics has become particularly associated with the University of Helsinki, whose VARIENG project (Nevalainen, Taavitsainen, and Leppa¨nen 1998–) continues to make a massive contribution to the historical study of English and related languages, particularly in the area of grammar, and with the University of Michigan, whose Middle English Compendium (McSparran [ed.] 2006) is a text-resource linked to the Middle English Dictionary (Kurath et al. 1952–2001). But the Helsinki and Michigan projects have tended to utilise critically edited rather than diplomatic texts, which – though often suitable for broad-brush syntactic studies, or for establishing the Middle English lexicon – have limitations when used for detailed study of, say, spelling practices. It is therefore no coincidence that the most significant study yet published of written language usage during the Middle English period, the Linguistic Atlas of Late Mediaeval English (LALME) (McIntosh et al. 1986), derived its information not from critical editions but largely directly from manuscript sources, supported where possible by input from diplomatic editions (an online and revised version of the Linguistic Atlas, e-LALME is forthcoming; see http://www.ling.ed.ac.uk/research/ihd/index.shtml). Perhaps the most impressive new direction in fresh diplomatic editing in relation to computational philology is The Canterbury Tales Project (Robinson and Bordalejo 1996–), which began as an attempt to develop a new kind of critical edition but which ended up with arguably a much more valuable resource: not only up-to-date transcriptions and bodies of associated apparatus, which have replaced the old Chaucer Society diplomatic editions produced by F. J. Furnivall and his Victorian associates, but also their presentation in machine-readable form. Because machine-readable, such editions allow for highly sophisticated text-mining, identifying spelling-systems and systems of punctuation which are of interest not just for linguists but also for book historians and students of changing literacy practices. Such online electronic corpora, based on diplomatic editing, are now an established part of linguistic enquiry, underpinning (e.g.) the Edinburgh Linguistic Atlas of Early Middle English (Laing and Lass 2007) and Linguistic Atlas of Older Scots (Williamson 2007), more recent Helsinki initiatives such as the online Corpus of Early English Correspondence (Nevalainen et al. 1998) and Corpus of Scottish Correspondence (Meurman-Solin 2007), or the Stavanger-Glasgow Middle English Grammar Corpus (Stenroos et al. 2011).
37
38
I Periods Diplomatic editions of Middle English texts are focused not so much on the author of the text as on the transmission of the text; i.e. they engage with scribes as informants. This focus on scribal output underpinned the development of the LALME, for instance. Researchers have showed that texts, when appropriately analysed, can be used as evidence for the language not of authors but of scribes, and, since scribes are, like authors, native users of Middle English, and of course comparatively much more numerous, the information they provide can be investigated for the purposes of linguistic (as opposed to literary) study.
3 Writing and speech The textual resources described at the end of the previous section are, and will continue to be, invaluable for the study of what might be termed the “transmission” of Middle English, viz. the systems of writing and speech. The mapping between writing and speech has been a concern of scholars since antiquity. In the classical and medieval west, authorities such as Donatus and Priscian adopted the “doctrine of littera”, which was developed to correlate speech with alphabetic writing-systems of the kind used for Greek or Latin. This doctrine distinguished between nomen (‘name [of the letter]’), potestas (‘power’ = ‘sound-value’), figura (‘representation’ = ‘written-symbol’), with the term littera (‘letter’) as a superordinate classificatory term. Although this ancient usage has been recently recuperated, most notably by scholars associated with LALME and its successor-projects, most students of Middle English use the accepted terminology to discuss transmission which has been developed in the 20th century. Broadly speaking, written languages are either “phonographic”, where there is a mapping (however conventional) between grapheme and phoneme, or “logographic”, where there is a mapping between a conventional symbol and a word or morpheme. The relationship between these different systems is of course clinal. Towards the logographic end is Chinese, whose conventionalized characters derive ultimately from pictorial representations of certain key concrete concepts, though this practice was rapidly modified to deal with more abstract notions: “Modern Chinese characters hold few really firm clues as to their pronunciation” (Newnham 1971: 44). Written Middle English, on the other hand, represents the opposite end of the cline, whereas Present-day English, with its various conventionalizations, is, while remaining broadly phonographic, rather closer to the logographic pole (see Rutkowska, Chapter 15). It is sometimes said that Middle English was written at a time when folk “wrote as they spoke”. Such a statement, of course, oversimplifies a complex situation; given that writing-systems are designed to give comparative permanence to something evanescent, i.e. speech, a degree of conventionalisation is to be expected. Nevertheless, the statement does summarise, albeit in broad terms, the phonographic status of Middle English. Given the historic primacy of speech over writing, it is no surprise that, until comparatively recently, the focus of scholarship has been on the reconstruction of Middle English speech-systems, i.e. the phonology of Middle English. Significant recent discussions include survey articles written for the relevant volumes of the Cambridge History of the English Language, especially the second volume (Blake 1992), and there are also useful introductory outlines in (e.g.) Horobin and Smith (2002) or Wright and Wright (1923). The most comprehensive outlines remain those by Luick (1964) and Jordan
3 Periods: Middle English (1974); however, both of these date in their essentials from the early 20th century, and both are seriously in need of updating. Despite this established scholarly tradition, there are approaches which have focused on the written mode as an object of enquiry in its own right. The best-known of these approaches is associated with the creation and exploitation of LALME and its successor projects, treating graphological features as objects of study without necessary reference to their spoken-language equivalents. A simple example makes the point. There is some evidence, supplied by LALME, that the form itt ‘it’ has a geographical distribution in Middle English; it is a dominant form in several texts localized to northern England, notably Cumberland. But it seems very doubtful that the -spelling signifies any sound-difference from . A more subtle difference, again identified as a result of work for the Linguistic Atlas (Benskin 1982), is to do with the allographic representation of the grapheme . (In this context, the term “grapheme” is used as the written language equivalent of the phoneme, i.e. the symbolic unit being aimed at by the scribe, while an “allograph” is the realisation of the grapheme in writing. Replacement of one grapheme by another changes the meaning of the word in which it occurs, but replacement of one allograph by another realisation of the same grapheme does not change the word’s meaning. It is conventional to place graphemes in angle brackets, thus: .) In general terms, in northern varieties of Middle English, seems to have been written identically with , whereas in southern varieties the two graphemes are realized distinctly. Again, there does not seem to be a sound-difference here, although the restriction of = in some northern texts to “function” words such as yat ‘that’, ye ‘the’ (cf. southern þat, þe) may correlate with the development of initial-fricative voicing in such cases. Another example of a spelling-usage which can be localized dialectally is initial in xal ‘shall’, xuld ‘should’, which is diagnostically East Anglian. How this feature maps onto pronunciation is again problematic; a plausible argument can be made that such forms represent a pronunciation-difference, but alternatively the usage could simply be a convention with only a local currency. The “ = ” example, of course, problematizes some of the conventional terminology used for the discussion of the writing/speech relationship. Briefly put, the problem is as follows: assignment of the allograph y to either the grapheme or the grapheme depends on the mapping of the allograph onto the sound-equivalent: is the form in question, in context, to be interpreted as a vowel or a semi-vowel, or is it to be interpreted as a fricative? Seeing y and þ as allographs of a grapheme raises some theoretical issues about the assignment of allographs to graphemes: can one allograph, in this case y, belong to two different graphemes? It is an axiom of phonological theory, of course, that by definition one allophone cannot belong to two different phonemes (although cf. neutralization of phonemes, for which see e.g. Gimson 1989: 50). Similar problems arise with (e.g.) and ; as is well-known, is used prototypically, in Middle English texts, to map onto either a vowel or a consonant when in medial position, whereas is used initially, e.g. vpon ‘upon’, loue ‘love’. Which mapping is made is determined by context. In Present-day English, the grapheme is similarly differentiated, mapping onto a consonant [j] in initial position but a vowel in final position, cf. yacht, dizzy. Michael Benskin (1982) simplifies this problem of categorization by recuperating the ancient doctrine of littera, and setting aside as unnecessarily complex the apparently
39
40
I Periods obvious parallelisms between grapheme/phoneme, allograph/allophone. According to Benskin’s approach, the figura , in some varieties, may be assumed simply to map onto two potestates, or sound-equivalents, with the nomen “thorn”. Developed discussions of these issues, using this terminology, have been adopted by others working in the LALME tradition (e.g. Laing and Lass 2009). Although it is true that there are many features of Middle English writing-systems which do not seem to map onto sound-differences, there are also features where such a mapping is much more likely. When a Northern scribe writes stane for ‘stone’ and a Southern or Midland scribe writes stone it seems likely that some sound-difference is being addressed; when a Southern scribe writes voules ‘birds’ (cf. PDE fowls) and a Northern or Midland scribe writes foules, again it seems likely that a sound-difference is being flagged. The issue, of course, is what these sound-differences really are. Our understanding of Middle English phonology is based on the interpretation of the following: a) “reconstruction”, both comparative (dealing with cognate languages) and internal (dealing with paradigmatic variation); b) analysis of “residualisms” surviving in modern accents of English; c) analysis of the writings of spelling-reformers and phoneticians from the Early Modern English period, supplying information about usages closer to the Middle English period than now; d) analysis of contemporary verse-practices, based on the analysis of rhyme, alliteration and meter; and e) analysis of spellings. None of these approaches, of course, provides direct evidence – there are no taperecordings from the Middle Ages – but they allow for the recuperation of the phonologies of Middle English in fairly broad terms. Details are given in standard handbooks on the subject, to which further reference should be made (e.g. Jordan 1974; see also Horobin and Smith 2002 and references there cited). Given the restrictions of space, an example might suffice to show not only what can be achieved using these various resources, but also the kind of problems faced by students of Middle English interested in reconstructing the sound-systems of the period. The example chosen here is from a 14th-century Middle English romance, Sir Orfeo. Sir Orfeo survives in three medieval manuscripts, of which the best known and the most authoritative is the Auchinleck Manuscript (Burnley and Wiggins [eds.] 2003); a facsimile, transcription and associated apparatus are available as an online resource. The Auchinleck Manuscript is a large miscellany in several hands; Sir Orfeo is now the thirty-eighth item contained within it. There is good evidence from external sources that the Auchinleck Manuscript was copied in London, but the rhyming practices in Sir Orfeo indicate that it was composed in the West Midlands. Evidence for such a localization includes “mistaken” rhymes such as frut ‘fruit’ (from Anglo-Norman fru¯t): lite ‘little’ (from OE ly¯t), man ‘man’ (OE mann): opan (OE uppon). The form opan seems to be an artificial formation, not attested elsewhere in Middle English and used here to sustain the eye-rhyme with the form man; forms for ‘upon’ all otherwise end in –on, and the rhyme would only work where ‘man’ appears as mon, prototypically a Western form. With the rhyme frut:
3 Periods: Middle English lite, the scribe has not bothered to sustain the eye-rhyme, and this too is suggestive. Old English ly¯t is reflected in lute, again, in Western dialects. A. J. Bliss, in his edition of Sir Orfeo, argues against a West Midland provenance for the poem on the basis of the form owy ‘away’ rhyming with cri ‘cry’ and fairy ‘fairy’; cf. mid-line oway elsewhere in the text; Bliss considers the form owy to be “unmistakably Kentish” (Bliss [ed.] 1966: lii). However, the Kentish associations of the form owy have been challenged (see comments by E. G. Stanley, cited by Bliss), and the remaining rhyming evidence, along with certain literary characteristics in the text, suggests a West Midland association. (See further Samuels 1955: 59.) The mapping of these rhymes onto speech can be reconstructed from an examination of residualisms in later accents of English. For instance, a rounded back vowel in ‘man’ was recorded in (inter alia) Warwickshire, Worcestershire, Shropshire, etc., by Joseph Wright, an excellent observer, at the beginning of the 20th century (see Wright 1905: 520); similar usages are noted in the mid-20th century (see e.g. Wakelin 1977: 96), where the form is seen as characteristic of the West Midlands. It therefore seems to be reasonable to suppose that this situation also applied in Middle English times, as suggested by spellings such as mon ‘man’. Such complex interpretative strategies as the one just illustrated exemplify the procedures used for reconstructing Middle English pronunciation; current developments in Middle English studies, most notably the completion of LALME and its successor projects, make, as indicated above, a comprehensive new interpretative survey a major scholarly desideratum.
4 Grammar Just as Middle English accents vary diatopically and diachronically, so do Middle English grammars. In broad terms, northern grammars tend to early innovation, while southern grammars tend to be more conservative, but this broad characterization misses the complex set of grammatical systems which existed in Middle English, and the various constraints which underpin them. Traditionally, the history of English grammar has been described in terms of the shift from synthesis to analysis, i.e. from a language which expresses the relationship between words by means of inflexional endings joined to lexical stems to one which maps such relationships by means of function-words such as prepositions. This broad characterization, of course, needs considerable nuancing, and can better be expressed as a comparatively short shift along a cline. Old English, in comparison with Present-day English, is comparatively synthetic, but nowhere near as synthetic as (say) non-Indo-European languages such as Present-day Finnish or Zulu, older Indo-European languages such as classical Latin – or even earlier manifestations of Germanic such as 4th-century Gothic, which, unlike OE, regularly distinguished nominative and accusative plural forms of the noun; cf. OE hla¯fas ‘loaves’ (both NOM and ACC), Go. hla´ibo¯s (NOM), hla´ibans (ACC). Present-day English is comparatively analytic, but not as analytic as (say) Present-day Mandarin Chinese; a 21st-century English-speaker still marks person, number and case, and sustains grammatical cohesion, with concord between verbal and pronominal inflexions, for instance, e.g. I love bananas beside she hates bananas. As might be expected, Middle English grammars occupy a half-way position between Old and Present-day English, although there is both diachronic and diatopic variation. Here in (1), for instance, is a short passage from one of the writings ascribed
41
42
I Periods to the Yorkshire writer Richard Rolle, as it survives in the Thornton Manuscript held in the Cathedral Chapter Library at Lincoln, and localized to Lincolnshire: (1)
Arestotill sais þat þe bees are feghtande agaynes hym þat will drawe þaire hony fra thayme. Swa sulde we do agaynes deuelis, þat afforces thame to reue fra vs þe hony of poure lyfe and of grace. For many are, þat neuer kane halde þe ordyre of lufe ynence þaire frendys, sybbe or fremmede. Bot outhire þay lufe þaym ouer mekill, settand thaire thoghte vnryghtwysely on thaym, or þay luf thayme ouer lyttill, yf þay doo noghte all as þey wolde till þame (Sisam 1921: 41) ‘Aristotle says that the bees are fighting against him who wishes to take their honey from them. So ought we to do against devils, who endeavor [literally, endeavor them, cf. the French reflexive verb s’afforcer] to take from us the honey of humble life and of grace. For there are many, who never know how to observe moderation in [literally, never know how to hold the order of] towards their friends, akin or not akin. But either they love them too much, focusing their mind unrighteously on them, or they love them too little’ (my translation)
Aspects of the spelling of this passage defamiliarize the text for a modern reader, e.g. swa ‘so’ and sulde ‘should’, both of which are distinctively northern in dialectal distribution (the former with northern as the reflex of Old English a¯, the latter with in ‘shall’, ‘should’ – demonstrating, incidentally, the similarity between northern Middle English and usages in West Germanic languages beyond the British Isles, cf. Presentday German and Dutch cognates). But in many grammatical features there are features which correlate, broadly, with Present-day English; thus verbs such as sais, are appear much as modern English, cf. PDE says, are, and the pronominal system is also similar to modern usage, e.g. þaire ‘their’, tha(y)me ‘them’, cf. OE hı¯e, him. There are some differences in the passage in inflexional morphology in comparison with Present-day Standard English. The form afforces, for instance, has an inflected form in –es despite being governed by a relative pronoun þat ‘who’ postmodifying a plural noun deuells ‘devils’, and the present participle ending –ande in feghtande might be compared with the Present-day Standard English -ing. However, such forms are only seen as deviant from the perspective of standard usage; -s-type endings in plural verbs, and a present participle inflexion in –an (derived from –ande) are both features of Present-day Scots, for instance. It is noticeable that, in the Rolle passage, the -s is dropped when the verb is immediately preceded by a third-person plural pronoun, e.g. þay luf(e); this structure, the so-called Northern Personal Pronoun Rule, is still a characteristic feature of Scots. (The origins of the Northern Personal Pronoun Rule have been much debated. Several scholars have suggested that the construction is derived through contact with a similar pattern in Celtic languages, and this view has recently been developed by Klemola 2000; Benskin 2011.) The Rolle passage might be compared with another 14th-century text from a quite distinct dialect area: The Ayenbite of Inwyt by Dan Michel of Northgate, which survives in a holograph manuscript (London, British Library Arundel 57), localized to Canterbury in Kent, in the extreme south of England, and dated to 1340. The following passage gives some flavor of the language of the text (2): (2)
Efterward þer wes a poure man, ase me zayþ, þet hedde ane cou; and yhyerde zigge of his preste ine his prechinge þet God zede in his spelle þet God wolde yelde an
3 Periods: Middle English hondreduald al þet me yeaue uor him. Þe guode man, mid þe rede of his wyue, yeaf his cou to his preste, þet wes riche. Þe prest his nom bleþeliche, and hise zente to þe oþren þet he hedde. Þo hit com to euen, þe guode mannes cou com hom to his house ase hi wes ywoned, and ledde mid hare alle þe prestes kun, al to an hondred. Þo þe guode man yseз þet, he þoзte þet þet wes þet word of þe Godspelle þet he hedde yyolde; and him hi weren yloked beuore his bissope aye þane prest. (Sisam 1921: 34–35) ‘Afterwards there was a poor man, as it is said to me, who had a cow; and [he] heard say from his priest in his preaching that God said in his gospel that God would repay a hundredfold all that one gave for him. The good man, according to the advice of his wife, gave his cow to his priest, who was rich. The priest took his [cow] happily, and sent her to the others that he had. When it came to evening, the good man’s cow came home to his house as she was accustomed, and led with her all the priest’s cattle, up to a hundred. When the good man saw that, he thought that that was the word of the Gospel that he had restored; and they were adjudged [literally, looked] to him before the bishop against the priest’ [i.e. the bishop ruled that a poor man should have the property in question, rather than the priest; see further Sisam 1921: 213]’ (my translation) Again, there are distinctive features of spelling in this passage, such as z, u/v where would appear in Present-day English, e.g. zayþ ‘says’, zente ‘sent’, uor ‘for’, wyue ‘wife’; this usage seems to reflect “Southern voicing”, still a feature of South-Western accents but recessive in Kent by Joseph Wright’s day (see Wright 1905: 226, 241). But in grammatical terms, the passage demonstrates southern retention of the older pronouns in h-, e.g. hi ‘they/she’, and even a relic of the inflected determiner, viz. þane ‘the/that’ (DAT). In phrases such as þe guode man ‘the good man’, adjectival inflexion is retained. In verbs, the Old English ge-prefix, which was apparently in origin an aspectual marker, is retained in reduced form as y-, in ywoned ‘accustomed’, yloked ‘adjudged’ etc., while the third person singular inflexion appears as –þ in the impersonal construction me zayþ ‘it is said to me’. In sum, the inflexional morphology of this passage is conservative in comparison with the northern usage represented by the Rolle text (see 1). Syntactically, Middle English retains some features characteristic of Old English which are no longer current in present-day usage. The following passage (3) is from a Western text, John of Trevisa’s translation of Higden’s Polychronicon, as it survives in London, British Library Cotton Tiberius D.vii, a manuscript dated to around 1400 and localized to Berkeley in Gloucestershire, on the English side of the Severn Estuary. (3)
Yn Britayn buþ meny wondres. Noþeles foure buþ most wonderfol. Þe furste ys at Pectoun. Þar bloweþ so strong a wynd out of þe chenes of þe eorþe þat hyt casteþ vp aзe cloþes þat me casteþ yn. Þe secunde ys at Stonhenge bysydes Salesbury. Þar gret stones and wondur huge buþ arered an hyз, as hyt were зates, so þat þar semeþ зates yset apon oþer зates. Noþeles hyt ys noзt clerlych yknowe noþer parceyuet houз and wharfore a buþ so arered and so wonderlych yhonged. (Sisam 1921: 146) ‘There are many wonders in Britain. Nonetheless four are most wonderful. The first is at “Pectoun” [a mistake for the Peak of Derbyshire; see Sisam 1921: 252]. There such a strong wind blows out of the fissures of the earth that it throws back up clothes that are thrown in [literally, that one throws in]. The
43
44
I Periods second is at Stonehenge besides Salisbury. There great and wonderfully huge stones are raised up high, in the manner of gates [literally, as it were gates], so that there seem to be gates set upon other gates. Nevertheless it is not clearly known or perceived how and wherefore they are raised up in this way and so wonderfully hung’ (my translation) The passage contains several features characteristic of a south-west Midland dialect, e.g. forms such as buþ ‘are’, with vowel in and plural inflexion in -þ, past participles with y-, e.g. yhonged, yknowe, or the form a ‘they’. But there are also syntactic archaisms; thus, the main verb is in “second position” after an initial adverbial, as in Old English and Present-day German, in constructions such as Þar bloweþ so strong a wynd out of þe chenes of þe eorþe, while gret stones and wondur huge ‘great and wonderfully huge stones’ is an example of the characteristically Old English construction known as the “splitting of heavy groups” (see Mitchell 1985: 612–616, Sections 1464–1471). The grammatical differences between the three passages analysed so far seem to derive from the different interaction each dialect has had with the language with which late Old English had had most intimate contact, viz. Old Norse. The impact of Norse on English was not really reflected in the written language in Anglo-Saxon times, since “classical” Late West Saxon Schriftsprache – based on the usage of the parts of England furthest from the Viking invasions – is by far the best recorded variety of English during the period, and was a language of official record (and thus liable to standardization) (see Kornexl, Chapter 24). Only during the Middle English period, with the loss of the national currency of the Schriftsprache, is the full impact of Norse to be discerned. It might be expected that the influence of Norse would be strongest in those areas where Viking settlement was most dense, viz. Northern England; and the fact that grammatical innovation is earliest and most thorough in northern varieties indicates that the impact of Norse encouraged the continuing shift, in English, from synthesis to analysis that characterizes – albeit with varying speeds and to varying extents – all the Germanic languages. Barbara Strang famously argued that the process “came first in the [North] because of its long exposure to bilingualism”: Those of us who have tried, as adults, to fit into a new language-community know that in such circumstances one does not retain any more of the morphology of the new language than is strictly necessary to make oneself understood. Where a community is unilingual, old patterns remain from inertia, regardless of their functional obsolescence. In this indirect way, far more than through direct syntactic borrowing, the presence of a Scandinavian community affected the syntax of English (Strang 1970: 281).
5 Lexicon Norse left its mark on the English lexicon as well as (it seems) its grammar. Most scholars hold that the third-person plural pronouns already cited as characteristic of Northern Middle English and now standard in Present-day English, viz. they, them, their have a Norse etymology, although their adoption may have been encouraged by congruence of the initial sound in Norse þeir, þeim, þeira with that of the so-called “simple” demonstrative of Old English, viz. þa¯, þæ¯m, þa¯ra, all of which could be used with pronominal function. A more complex example is the form she, which seems to demonstrate the
3 Periods: Middle English impact of Norse in a more roundabout way. Most scholars hold that this form arose from a blend of the Old English he¯o with a Norse pronunciation, yielding *hjo¯; such forms seem to be represented in spelling by (e.g.) ʒho. The /hj/-cluster, being of low yield in the English lexicon, has a tendency to shift to [ ʃ ], cf. the form Shug, a modern nickname for the common Scottish personal name Hugh, or the commonly-cited example of Shetland from older Hjaltland (see further Britton 1991; see also Samuels 1972: 114–116). Adoption of the Norse forms of the pronouns seems to have been encouraged by general inflexional loss; more distinctive pronouns were required as syntactic tracking devices, and it is noticeable that the Present-day English third-person pronouns he, she, it and they are much more phonetically distinctive than the Old English equivalents he¯, he¯o, hit and hı¯e. It is probably not a coincidence, therefore, that it was in Kent, where inflexional distinctions were best preserved, that the new distinctive pronouns were slowest to develop – although of course this slowness would have been reinforced by the fact that the new forms, because derived from Norse, were not available in that part of the country for early selection. Norse also left its mark on the set of open-class words commonly occurring in English; common words such as ill, wing, egg, take, skill are all from Norse. Some words derived from Norse but with Old English cognates have, interestingly, developed distinct meanings, e.g. Norse-derived skirt beside Old English-derived shirt. Because of the closeness of the relationship between the two languages – particularly between Norse and the most northerly dialect of Old English, Old Anglian – it is occasionally hard to work out whether the Present-day English word is derived from Norse or from a dialect of English; a good example of such an uncertain form is call, which is probably derived from Norse kalla but is also recorded once in Old English as ceallian (see the etymology in OED for call v.). Norse-derived place-names often indicate Norse settlement, even if the forms involved are no longer part of the local dialect; thus Ormskirk in Lancashire contains the Norse generic kirk ‘church’, flagging a time when the local word for ‘church’ was kirk. However, the form is recessive, and is now largely restricted to Scottish use, referring to the established Church of Scotland (see Volume 2, Dance, Chapter 110). Whereas Norse-derived words seem to fall for the most part into everyday domains of language during the Middle English period, other words taken from other languages are of more restricted currency. Into this group fall what are by far the largest body of new items entering the English lexicon, viz. words derived from varieties of French. The earliest French-derived forms which seem to have become part of the English lexicon were taken, as might be expected, from Norman French, e.g. prison, grace, mercy, baron etc.; later on during the Middle English period many words are taken from Central French. Sometimes an early loan from Norman French can be distinguished in English by its pronunciation; thus Present-day English war comes from Norman French werre rather than from Central French guerre. Sometimes Norman and Central French cognates have both entered the language, cf. warden alongside guardian. Many French stems have been combined with English affixes to produce new words, e.g. gentleness (see Volume 2, Skaffari, Chapter 106). The influx of French words also had a structural effect on the semantics of the English lexicon: new words took over slots hitherto occupied by native forms, and a reorganisation followed. For instance, the French-derived word memory, first recorded in
45
46
I Periods English in the 13th century, took over from Old English (ge)mynd, which in turn replaced the obsolete and largely poetical form hyge with the meaning ‘mind’ (see Smith 1996: 138). The study of the shifting semantic structure of English is currently in its infancy. The recent completion of the Historical Thesaurus of the Oxford English Dictionary (Kay et al. 2009) will allow for a major leap forward in our understanding of semantic change with reference to the lexicon; the possibilities for such work have already been demonstrated by the online Thesaurus of Old English (Edmonds et al. 2005; also Roberts et al. 2000). The primary resources for the study of the Middle English lexicon remain, however, not only the Oxford English Dictionary (Simpson [ed.] 2000–), on which the Historical Thesaurus draws and with which it is now linked online, but also the Middle English Dictionary (Kurath et al. 1952–2001). The Middle English Dictionary, whose first fascicule appeared in 1952 but which derived originally from William Craigie’s call, back in 1919, for a series of “period” dictionaries to complement the Oxford English Dictionary, was completed in print form in 2001. It was transferred online shortly afterwards as part of the larger Middle English Compendium (McSparran [ed.] 2006; http://quod. lib.umich.edu/m/mec/digitMSS.html) which also includes a corpus of machine-readable texts and a hyper-bibliography, greatly increasing its functionality. The online version of the Middle English Dictionary makes it possible for complex searches to be performed; it also enhances massively our understanding of the range of variation that existed in Middle English.
6 Summary It will be clear from the above that the study of Middle English is at an exciting stage. Middle English is, for the reasons given above, the period in the history of English when variation in the spoken mode is reflected (however partially) most thoroughly in writing, and the variationist paradigms now dominant in theoretical studies offer scholars many exciting avenues for future research. The combination of the new resources now available alongside major developments in linguistic theory means that a reconciliation of “old” philology and “new” linguistics is now possible (see the very apposite comments in Rissanen 1990, which prefigured many current developments in the field).
7 References 7.1 Online resources Burnley, David and Alison Wiggins (eds.). 2003. Auchinleck Manuscript. National Library of Scotland. See: http://www.nls.uk/auchinleck/ Edmonds, Flora, Christian Kay, Jane Roberts, and Irene´ Wotherspoon. 2005. Thesaurus of Old English. University of Glasgow. http://libra.englang.arts.gla.ac.uk/oethesaurus/ Kurath, Hans, Sherman M. Kuhn, John Reidy, and Robert E. Lewis. 1952–2001. Middle English Dictionary. Ann Arbor: University of Michigan Press. http://quod.lib.umich.edu/m/med/ Laing, Margaret and Roger Lass. 2007. LAEME: A Linguistic Atlas of Early Middle English, 1150–1325. University of Edinburgh. http://www.lel.ed.ac.uk/ihd/laeme1/laeme1.html McSparran, Frances (ed.). 2006. Middle English Compendium. Ann Arbor: University of Michigan Press. http://quod.lib.umich.edu/m/mec/. See also http://quod.lib.umich.edu/m/mec/digitMSS.html
3 Periods: Middle English Meurman-Solin, Anneli. 2007. Corpus of Scottish Correspondence. University of Helsinki. http:// www.helsinki.fi/varieng/CoRD/corpora/CSC/; for manual, see: http://www.helsinki.fi/varieng/ csc/manual/ Nevalainen, Terttu, Irma Taavitsainen, and Sirpa Leppa¨nen. 1998– The Research Unit for Variation, Contacts and Change in English (VARIENG). Department of English, University of Helsinki. http://www.helsinki.fi/varieng/index.html Nevalainen, Terttu, Helena Raumolin-Brunberg, Jukka Kera¨nen, Minna Nevala, Arja Nurmi, and Minna Palander-Collin. 1998. Corpus of Early English Correspondence (CEEC). Department of English, University of Helsinki. http://www.helsinki.fi/varieng/CoRD/corpora/CEEC/index.html Robinson, Peter and Barbara Bordalejo. 1996–. The Canterbury Tales Project. Institute of Textual Scholarship and Electronic Editing, University of Birmingham. See: http://www. canterburytalesproject.org/ Simpson, John (ed.). 2000–. The Oxford English Dictionary. 3rd edn. online. Oxford University Press. www.oed.com/ Stenroos, Merja, Martti Ma¨kinen, Simon Horobin, and Jeremy Smith. Version 2011.2. The Middle English Grammar Corpus (MEG-C). http://www.uis.no/research/culture/the_middle_english_ grammar_project/meg-c/ Williamson, Keith. 2007. A Linguistic Atlas of Older Scots, Phase 1: 1380-1500 (LAOS). http:// www.lel.ed.ac.uk/ihd/laos1/laos1.html
7.2 Printed resources Benskin, Michael. 1982. The letter and in later Middle English, and some related matters. Journal of the Society of Archivists 7(1): 13–30. Benskin, Michael. 2011. Present indicative plural concord in Brittonic and Early English. Transactions of the Philological Society (Special Issue on Languages of Early Britain, ed. by Stephen Laker and Paul Russell) 109(2): 158–185. Blake, Norman (ed.). 1992. The Cambridge History of the English Language. Vol. II: 1066–1476. Cambridge: Cambridge University Press. Bliss, A. J. (ed.). 1966. Sir Orfeo. 2nd edn. Oxford: Clarendon Press. Britton, Derek. 1991. On ME she/sho: A Scots solution to an English problem. NOWELE 17: 3–51. Dobson, Eric J. (ed.). 1972. The English Text of the Ancrene Riwle: The Cleopatra Text. Oxford: Early English Text Society. Gimson, A. C. 1989. An Introduction to the Pronunciation of English. London: Arnold. Horobin, Simon and Jeremy Smith. 2002. An Introduction to Middle English. Edinburgh: Edinburgh University Press. Jordan, Richard. 1974. A Handbook of Middle English Grammar: Phonology. Rev. and trans. by Eugene Crook. The Hague: Mouton. Kane, George and E. Talbot Donaldson (eds.). 1974. Piers Plowman: The B-Text. London: Athlone Press. Kay, Christian, Jane Roberts, Michael Samuels, and Irene´ Wotherspoon (eds.). 2009. Historical Thesaurus of the Oxford English Dictionary. Oxford: Oxford University Press. Klemola, Juhani. 2000. The origins of the northern subject rule: A case of early contact? In: Hildegard L. C. Tristram (ed.), The Celtic Englishes. Vol. II, 329–346. Heidelberg: Winter. Laing, Margaret and Roger Lass. 2009. Shape-shifting, sound change and the genesis of prodigal writing systems. English Language and Linguistics 13(1): 1–31. Lass, Roger. 2006. Historical Linguistics and Language Change. Cambridge: Cambridge University Press. Luick, Karl. 1964. Historische Grammatik der englischen Sprache. Oxford: Blackwell. McIntosh, Angus, Michael Samuels, and Michael Benskin, with Margaret Laing and Keith Williamson. 1986.ALinguisticAtlasofLateMediaevalEnglish(LALME).Aberdeen:AberdeenUniversityPress. McMahon, April. 1994. Understanding Language Change. Cambridge: Cambridge University Press.
47
48
I Periods Mitchell, Bruce. 1985. Old English Syntax. Oxford: Clarendon Press. Newnham, Richard. 1971. About Chinese. Harmondsworth: Penguin. Reynolds, Leighton and Nigel Wilson. 1974. Scribes and Scholars: A Guide to the Transmission of Greek and Latin Literature. Oxford: Clarendon Press. Rissanen, Matti. 1990. On the happy reunion of English philology and historical linguistics. In: Jacek Fisiak (ed.), Historical Linguistics and Philology, 353–370. Berlin/New York: Mouton de Gruyter. Roberts, Jane and Christian Kay with Lynne Grundy. 2000. A Thesaurus of Old English in Two Volumes. 2nd revised edn. Amsterdam/Atlanta: Rodopi. Samuels, Michael. 1955. Review of Sir Orfeo, ed. by Alan J. Bliss. Medium Aevum 24: 56–60. Samuels, Michael. 1972. Linguistic Evolution. Cambridge: Cambridge University Press. Sisam, Kenneth (ed.). 1921. Fourteenth-Century Verse and Prose. Oxford: Clarendon Press. Smith, Jeremy. 1996. An Historical Study of English. London: Routledge. Strang, Barbara. 1970. A History of English. London: Methuen. Wakelin, Martyn. 1977. English Dialects. London: Athlone Press. Wright, Joseph. 1905. An English Dialect Grammar. Oxford: Frowde. Wright, Joseph and Elizabeth M. Wright. 1923. An Elementary Middle English Grammar. Oxford: Clarendon Press.
Jeremy J. Smith, Glasgow (UK)
4 Periods: Early Modern English 1 2 3 4 5 6
Introduction Historical and social background Printing and vernacularization Resources for the study of Early Modern English Changes in Early Modern English References
Abstract The two most notable changes in the Early Modern English period (1500–1700) were standardization and the growth of the lexicon. Changes in the cultural and political climate, such as the spread of printing and increasing availability of education and subsequent growing literacy among the population, were linked to these changes. The process of vernacularization in many areas (science, religion, law, government) produced new uses for English, and the Renaissance ideals of writing produced new styles and registers. Increased mobility, particularly towards London, contributed to the spread of linguistic innovations. The progressively more global trade brought contacts with new languages, and the spread of English world wide took its first steps in the colonialization of North America.
1 Introduction Early Modern English is perhaps most commonly said to range from 1500 to 1700, but since language change is gradual rather than abrupt, such demarcation lines are Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 48–63
4 Periods: Early Modern English naturally abstractions. Late Middle English during the 15th century increasingly shows features typical of Modern English, becoming more easily understandable even to the untrained present-day reader. The spread of the printing presses, one of the shaping forces on the development of Early Modern English, started towards the end of that century. Similarly, the cut-off date at 1700 marks the approximate time when most great changes during the Early Modern English period had run their course, and leaves the heyday of the prescriptive and normative tradition of the 18th century outside this stage in the development of the language. The social, political, and cultural changes associated with the Renaissance all influenced the development of the English language in the early modern period. These changes were in many ways interrelated, and reinforced each other. The preference for studying classical sources instead of the medieval authorities’ commentaries on them (Ad fontes) led to an educational reform, benefiting from the new appreciation of learning. The new schools provided literacy for an increasingly large part of the population. The changes in the intellectual climate and educational opportunities were both tied to the advancement of science and the concomitant vernacularization process of scientific writing. The Reformation, with a gradual break from the Latin traditions of the Catholic Church and the vernacularization of religious life, showed a parallel trend in stressing the authority of the original source, the Bible, and the need for people to have the ability to study it firsthand. Similar trends of vernacularization can be found in other areas of life, such as politics and law. Finally, the Age of Discovery provided contacts with new cultures as well as the beginnings of colonialization and, on the linguistic front, the first stages of American English. Many changes had a direct impact on the daily lives of the population, and the printing press was instrumental in disseminating these trends to the reading public. On the level of linguistic change, the two most notable processes are the standardization of written language and the vast increase in the lexicon. Much attention has been paid to the standardization process, which tends to provide an overly narrow view of the language as a whole, since dialects continued to be spoken (and in some cases written) by a vast majority of the population, even if this is disguised in the evidence remaining to us. While standardization of particularly printed sources tends to mask existing linguistic variation, there are also sources which give us a new perspective on the language. From the point of view of the linguist studying the period, the most important difference with earlier centuries is the wealth of new evidence on the linguistic practices of the population, providing us with English that was never written down before, or not in such quantity. Not only are there more types of texts (such as scientific and religious writing) being written in English, there are also more people than ever leaving written evidence concerning their lives. “Ego documents”, such as letters, autobiographies, wills, and travelogues, all have first person singular in common, but they also all give first-hand evidence of the linguistic practices of people in often quite private and personal, informal circumstances. We are still at the mercy of what has been preserved, but because more texts of all kinds were produced, there are also more kinds of writing remaining. There are few general descriptions of Early Modern English. The most extensive of these is the third volume of the Cambridge History of English (Lass 1999). There are also three book-length introductions aimed primarily at an undergraduate readership: Barber (1997), Go¨rlach (1991), and Nevalainen (2006). Each represents a particular
49
50
I Periods stage in the scholarship on Early Modern English, as can be most clearly seen in the evidence they use to support their description of the language. Barber illustrates his volume mostly through literary texts, following the tradition of stressing the importance of Shakespeare, and the value of drama as evidence of spoken language. Go¨rlach includes an ample selection of texts from a number of genres, bringing a wider perspective by the inclusion of more formal, non-literary types of writing. Nevalainen is a representative of the present-day paradigm of corpus-based research and draws her examples from electronic corpora and databases (see Section 4 below). All approaches have their merits, and together they provide a fuller picture of what the English language was like. Together they also illustrate the varied approaches it is possible to take when studying the language of the early modern period. This chapter has its main focus on the social, cultural, and political contexts in which Early Modern English was produced (Section 2) and on the production of English language texts (Section 3). A brief introduction to resources for the study of Early Modern English is provided in Section 4. The last section gives a brief overview of some of the changes in the language of the period, but leaves the more detailed discussion of all the linguistic aspects of Early Modern English to be found in the relevant chapters elsewhere in these volumes.
2 Historical and social background Language variation and change never take place in isolation. The connections between language and society mean that historical events need to be taken into account also when discussing the overall developments of Early Modern English. While Lass (1999: 5) is certainly correct in asserting that “[t]he story of a language ‘itself’ must be carefully distinguished from the story of its changing uses, users and social context – just as the changes themselves (as results) must be distinguished from the mechanisms by which they came about (e.g. lexical and social diffusion)”, it would still be remiss of us to overlook the influence of political and social changes taking place in the society where Early Modern English was being spoken and written. These changes required language users to adapt to new situations by creating new words and new styles of writing, they facilitated or hindered the dissemination of changes, and they influenced the variety of texts produced and preserved to us for study. Arguably, the most significant political events in the history of the period (at least when looked at from a linguistic point of view) were probably the Reformation and the consequent dissolution of the monasteries in the 16th century and the Civil War in the 17th century. Both episodes led to increasingly loose network ties because of the increased mobility of the population, and these in turn sped up linguistic change for some variables (Milroy 1992; Raumolin-Brunberg 1998). Both also produced, directly or indirectly, new types of texts, which allowed English to be written down in ways unlike those of previous centuries. The main development in social history relevant for language change was the advancement of educational opportunities for a wider range of social strata and the subsequent increased literacy rate of the population. This, in turn, gives us more linguistic first-hand evidence from a larger proportion of the population. The more widespread literacy also tied in with the advancement of printing, since there was a more extensive reading public than before. The availability of new genres for a lay audience and the
4 Periods: Early Modern English growing vernacularization of genres such as science and religion, which had previously been mainly the province of Latin, were all part of the larger picture. The population of England increased rapidly during the 16th century, followed by a time of stagnation before a further increase began in the 18th century (Coleman and Salt 1992: 2). Since there were no reliable statistics or census data created at the time, estimates of population size have been made based on such divergent data as muster rolls, lay subsidy rolls, ecclesiastical censuses, and parish registers. The reliability of population information increases when “a modernizing mercantilist state” required accurate information of resources and security, but also increasing literacy and numeracy and a more settled social and political order contributed to the development (Coleman and Salt 1992: 7). A summary of different population estimates suggests that the mid-16th-century population of England was somewhere around 2.8 million, rising to 4.1 million in the early 17th century, and showing a reasonably steady 5.0–5.2 million in the latter half of the 17th century (Coleman and Salt 1992: 5). The population of Scotland around 1600 has been estimated at one million, stagnating after that, while Ireland went from the same one-million population in 1600 to twice that in 1700; the North American English-speaking population started from a few thousand and reached quarter of a million by 1700 (Kishlansky 1996: 8). The population of London increased at a much more rapid pace than that of the country in general, from 50,000 in 1500 to 200,000 in 1600 and 575,000 in 1700 (Coleman and Salt 1992: 28). This shows the growing importance of the metropolis as a hub of government, commerce, and culture, and points to a special position also from the point of view of linguistic development (see Nevalainen, Volume 2, Historical Sociolinguistics). Already in medieval Britain there was a fair amount of geographical mobility and a later age of marriage than seems to have been common elsewhere in Europe. This was due to a free market in land, labor, and food. During the early modern period the average household consisted of 4.7 people, who were members of the nuclear family, and only rarely were there three generations under one roof. Up to 30% of households included servants, i.e. resident household or farm workers, who were typically single, aged between 15 and 30, and both men and women. As many as three quarters of boys and half of girls were in service at some point of their life. The common practice of service increased the geographical mobility of the population even further. The average age of first marriage in the 17th century for men was around 28 and for women 26 (Coleman and Salt 1992: 7, 14–15). This pattern of population movement had an impact on linguistic change in promoting dialect contact. As the population increased, there was even greater pressure for migration. While there was a great deal of subsistence migration by the unskilled and poor, particularly after 1650 there was also a large number of skilled people migrating to better themselves. Especially Scotland and Ireland produced a constant stream of migrants both to England and later to the New World (Kishlansky 1996: 13). We can only speculate how different the linguistic patterns of the mostly unlettered subsistence migrants and the at least minimally educated skilled migrants were, and how much influence either group would have had in the new location they settled at. The effect of social ambition on linguistic patterns has been established, so it is plausible that skilled workers would have more resources for linguistic adaptation. The major population crises during the early modern period were epidemics of the sweating sickness in the mid-16th century and recurring plague. In London, over 15%
51
52
I Periods of all deaths between 1580 and 1650 were caused by the plague. There were significant epidemics in 1563, 1593, and 1603, with minor outbreaks in 1578 and 1582 (Rappaport 1989: 72). The epidemics increased mobility among the population in two ways. On the one hand, the number of deaths meant there was more room for newcomers from all over the country. On the other hand, people fleeing London because of the danger probably took their new city ways, including any linguistic innovations, with them. In the 17th century, the Civil War had its own cost in loss of life: the estimated number of deaths was 80,000 in combat and 100,000 from disease (Coleman and Salt 1992: 24). Overall, the death rate in towns was higher than the birth rate, which made constant migration necessary for their growth. The continuous stream of migrants to London meant that English from all over the country could be heard in the streets, even if some areas (particularly North England) were over-represented (Coleman and Salt 1992: 27). During the early modern period, unlike the medieval times, there was no major, linguistically significant, influx of immigrants from abroad, but there were foreign craftsmen who moved to England to stay. Many of these were Protestants escaping religious persecution on the continent, but also skilled craftsmen from various countries seeking a livelihood. Around 1500 one in ten craftsmen in London were immigrants, and by 1540 they numbered one in six. In the rest of the country they were found in smaller numbers (Youings 1984: 128). Again, we can only speculate on the influence of these people, but arguably they would have had some influence on the professional language of their particular trade if nothing else. In contrast to earlier periods, there was more emigration from England, Ireland, and Scotland. The colonization of North America began, and after the first wave of migrants the surplus population of Ireland and Scotland was over-represented among those seeking opportunities in the New World. This obviously had an impact on how the new variety began to be shaped. As a result of new trade routes being discovered, English merchants, such as the East India Company, could be found trading at far distant places from the late 16th century onwards. While the original intent was to trade mostly with the East Indies, the company founded trading posts all over the Asian coastline, including India and China, but also e.g. Japan. There were also other trading companies in the West Indies and West Africa, but none was quite as long-lived or influential. The trading contacts led not only to new vocabulary for previously unknown peoples, cultures, and merchandise, but also to yet another new genre of popular writing, the travelogue. Contact was maintained not only with the indigenous peoples but also with other European traders working in the same areas, which led to the creation of trading jargons, and numerous letters were sent home, describing conditions of trade.
2.1 Education The humanist ideas of the Renaissance led to an increased appreciation of education. The aristocracy began to maintain the ideal of the well-rounded gentleman, which included learning. At the same time, education was increasingly seen as the means of providing the country with competent public servants. This was a trend that had already started in the 15th century, but it became increasingly important in the early modern period. Since the reorganization of the Tudor state and the expansion of government
4 Periods: Early Modern English activities, as well as the increase in diplomacy and foreign trade, came with a concomitant need for voluminous correspondence and detailed record keeping, there was a constant demand for literate and learned civil servants (Briggs 1994: 97; Cressy 1975: 5). Education was increasingly a secular business (Youings 1984: 17). While the dissolution of the monasteries had led to the end of schools in connection with monasteries, it provided the country with unemployed monks, who were often able to work as freelance school teachers in informal schools over the country. At the same time, new secular schools were being founded. Henry VIII alone is linked to at least eighteen schools founded or re-established during his reign, and many of the nobility followed suit. Schoolmasters needed a license to teach, and before the Civil War these licenses were under ecclesiastical control, with the purpose of preventing Catholic and Puritan teaching. During the Commonwealth control was shifted to the Parliament, but the success of any authority on imposing their demands on individual teachers is likely to have been limited at best. Education, like everything else in the society, depended on social status. The number of schools increased by at least 300 in 1500–1620, but the type of schools was extremely varied. At the one end, there were small private schools kept by a single master, at the other, grammar schools with wealthy patrons (Briggs 1994: 123). In 1647, educational reformer Samuel Hartlib envisioned four different types of education for the different social strata: one for the “vulgar, whose life is mechanical”, another for the gentry, “who are to bear charges in the commonwealth”. The third kind of school should be for scholars, who would go on to be teachers, and the fourth for the ministry (cited from Cressy 1975: 23). For the highest ranks of society, education was a value in itself, often initially received at home from tutors, while the lower ranks saw education as a means of social advancement, and were more typically educated at the various types of schools (Youings 1984: 119–120). Apart from social rank, gender was another major influence on the type and breadth of education available for individual people. While there were exceptional women at any given time, highly educated and well-read, they were definitely in the minority. Women like Margaret Roper, Katherine Astley, Queen Elizabeth, Ann Conway, or Dorothy Osborne were the exception, not the rule. Formal education was mostly unavailable, although some girls attended petty schools, small elementary schools often run by a single teacher. With higher social status came the possibility of private tutoring, at times including even classical learning. Since women did not work outside the home in professions where literacy and learning were needed, teaching them anything beyond basic skills was not considered a priority. In 1581 Richard Mulcaster, when discussing the education of women, suggested that “[r]eading if for nothing else (…) is very needful for religion, to read that which they must know and ought to perform” (cited from Cressy 1975: 110). A century later, in 1673, Bathusa Makin started a private academy for young ladies at Tottenham, but was very conscious of the resistance she was likely to meet: “I expect to meet with many scoffs and taunts from inconsiderate and illiterate men, that prize their own lusts and pleasure more than your [= that of ‘all ingenious and virtuous ladies’] profit and content” (cited from Cressy 1975: 113). Increased educational opportunities appeared at all levels of schooling: the number of students attending universities rose, and new groups of people, such as parish priests, had a university degree. For lay people, universities were a means of social advancement in the administration, but many depended on a wealthy patron or a scholarship
53
54
I Periods to pay for their expensive education. Legal training at the Inns of Court was almost entirely beyond people below the rank of gentry because of the prohibitive cost (Briggs 1994: 124). In the end, sources describing the realities of education leave much for conjecture, and a great deal of what we know is based on estimations, but there is a body of writing on the theory. A great many handbooks describing the kind of education that was desirable were published, ranging from Elyot’s Book Named the Governor (1531) to Ascham’s Schoolmaster (1570) and Mulcaster’s Positions (1581). What the existence of these books shows us is that the content of a suitable education was in general agreed upon. The handbooks range from the philosophical to the practical, discussing the education of all strata of society (Brink 2010: 31).
2.2 Literacy In the beginning of the early modern period literacy in England was restricted to the elites of society, but the transition to mass literacy began during this time (Cressy 1980: 175). Around 1500, the estimated rate of illiteracy for men was approximately 90%, while for women it was still very close to 100%. The literate people belonged to the highest strata of the population, and literacy was, for talented young men lucky enough to find a patron to support their education, a way of social advancement. Around 1600, illiteracy had clearly decreased, close to 30% of men being literate, but still only 10% of women. By 1700, the change is remarkable, since nearly half of all men could read and write and a quarter of women as well (Cressy 1980: 177). It should be remembered that these statistics are based on estimates, and that some of the assumptions at the basis of them are not completely reliable. It may well be, for example, that a person may sign a document with a mark and yet be able to write. Particularly the literacy of gentlewomen may be underestimated. There were many occasions when it was vitally important for a woman de facto looking after the estates in the absence of her husband to be able to keep account of household matters, to oversee the work of scribes employed by the family and the like, and this could be achieved more reliably if the mistress of the house was herself literate. At the same time, when scribes were available, women may simply have preferred to make use of their services rather than writing themselves (Brink 2010: 28–29). Literacy in early modern England was taught as two separate skills, reading and writing. This means that those who could read were not necessarily able to write. Reading was advocated by religious and secular writers alike. The ability to read the Bible was considered to be a spiritual benefit of great value to the general public. At the same time, education was seen as having both a moral and a civic value (Cressy 1980: 186). As mentioned above, different strata of the population did not have equal access to education, which also leads to literacy being unequally represented among them. It should be remembered, however, that literacy was not necessarily learned at school, but could also be taught by a family member or employer. Boys apprenticed to craftsmen and merchants were usually expected to have an elementary command of literacy and numeracy (Youings 1984). They would then be further instructed by their masters in the skills specific to their trade. Letter writing, for example, was often learned by copying old letters. This transferred not only the spelling conventions of the writing community, but also the textual practices involved in that particular genre. In addition
4 Periods: Early Modern English to factual literacy, being able to transfer one’s thoughts to paper, it was often necessary to be familiar with genre conventions and the requirements of a particular author-audience relationship. While some social conventions of spoken language (such as forms of address) could be more or less directly transferred to written form, there were other practices in the areas of, for example, style shifting and deferential discourse which were probably more tied to the written expression of social relations. In addition, much depended on a writer’s command of the “rules” of written language (see e.g. Palander-Collin 2009).
3 Printing and vernacularization One of the greatest changes in the early modern period when compared to the Middle Ages was the proliferation of all kinds of writing in English. As mentioned above, this was linked to the cultural developments related to the Renaissance and to the greater number of literate readers as a lay audience for new genres. Both entertainment and information of all kinds were reaching its readership. This new proliferation of different types of texts was possible because there was a reading public willing to pay. The mass-production of books, the increased literacy, and the relative affluence of middle ranks gave rise to a new audience for the more popular sorts of writing. Fiction of all kinds – prose, verse, and drama – was published in increasing amounts, ranging from broadside ballads to multivolume collections of plays. Pamphlets were produced to enlighten the public, to present political views, and to introduce new ideas and inventions. For example, sermons, the pros and cons of tobacco smoking, and new scientific discoveries were all topics suited for this form of publication. Handbooks providing instruction on many fields from medicine and culinary recipes to letter writing and proper conduct in polite society were increasingly made available to the lay readership. For the linguist, there is also much more surviving data than from earlier centuries, probably because so much more was written, both for publication and for private audiences. Many examples of private writing, particularly ego documents such as private letters and journals, remain unedited in archives, but the wealth of edited data is significantly more varied than in previous centuries. On the one hand, English was being used in new kinds of written language, presenting us with registers which either did not exist in earlier centuries or were curtailed to spoken language. On the other hand, because of the increased proportion of literate people, we are gaining direct access to the language of an ever widening part of the population. We are no longer solely reliant on fiction for the language of the middle and even lower ranks, since they – or at least some of them – are able to put pen to paper themselves. Similarly, women’s voices are more clearly heard during the early modern period than ever before. As the legal system increasingly functioned in English, court proceedings were also beginning to be recorded (and published) in that language, giving further voice to the previously silent. This means that our understanding of the full range of English in use is more complete than during earlier centuries. We are still far from actual spoken language, but we are getting a better idea of private and informal language from the actual speakers themselves. Personal correspondence is one obvious genre, made necessary, for example, by the mobility of people, as they entered service, moved to London to find their fortunes, or married outside their own immediate locality. There are also more personal journals,
55
56
I Periods commonplace books, and household accounts, which all reveal the more private and often informal side of people. On the more official side, the number of documents prepared by the growing number of civil servants increases notably during the period, and these documents are more typically written in English than during the Middle Ages, when they were more often written in Latin and Norman French. Because we have very little corresponding material from earlier periods, it is often difficult to estimate whether some words or forms of expression are new to the age or have simply never been written down before – or at least not in a form that has survived to us.
3.1 Printing Printing was a way of disseminating ideas, but also a way of disseminating the emerging written standard language. Printed books had a wider circulation than manuscripts (and many genres still circulated largely in that form), but it is notoriously difficult to estimate how great a difference this made to the actual size of the reading public. Ownership of books was certainly fashionable, and a way of displaying wealth (Youings 1984: 194). The number of books printed each year increased steadily, and in addition to books, there were pamphlets and broadsheets (Briggs 1994: 123). The book trade had its centre in London, but was by no means confined to it. Major towns had their established book sellers, and books were available at markets and fairs, by traveling peddlers along with other merchandise. During the third quarter of the 16th century nearly 4000 books were published, and during the last quarter this nearly doubled. In the 17th century, nearly ten thousand books were published in each quarter century (O’Callaghan 2010: 165). One of the consequences of the educational system becoming more regulated was a greater degree of shared background amongst the educated, and a widening of areas of interest. The ideal of the “Renaissance Man” included both literary culture and the visual arts, but also physical skills such as fencing, shooting, riding, and dancing (Briggs 1994: 124). This led to an interest in guides and handbooks in the various areas of expertise deemed necessary for the perfect courtier. Also other books, ostensibly aimed at a more common readership, were in actual fact aimed at the highest ranks. For example, Fitzherbert’s Book of Husbandry, published in 1523, was quite expensive and had a small print run of a few hundred copies (Youings 1984). There was also an increasing interest in news, which led to the publication of newsbooks from the early 1620s onwards. There were newsletters that readers could subscribe to, and these were often distributed in manuscript form, but printing was eventually the way for news as well. By the end of the 17th century, there were numerous news sheets being published, as well as twice- or thrice-weekly newspapers such as the London Gazette or the British Mercury (Briggs 1994: 165–166). Cheap, popular writing of the era included ballads, chapbooks, almanacs, and jestbooks, as well as other types of fairly ephemeral writing, which has often been regarded as representing the literary tastes of the lower ranks of society (Barry 1995: 73). On the other hand, escapist literature in the form of chivalric romances continued to be quite popular, and it is more than likely that the readership of these books went far beyond the highest strata of literate society (Barry 1995: 74). Texts were translated, abridged, rewritten, and sampled for the benefit and pleasure of those not able to read them in the original (Barry 1995: 80). Snippets were published as unbound books and in
4 Periods: Early Modern English newspapers and magazines, which made them available at a lower price. Texts would be shared by several people by reading aloud in places where people gathered, which further lowered the cost for each reader (Barry 1995: 81).
3.2 Vernacularization Early modern England was no longer a multilingual country in the way medieval England was. While Celtic languages continued to be spoken in the west, the Norman French aristocracy had seemingly lost their language by the 16th century. Despite the loss of societal multilingualism, functional multilingualism continued in many ways. Latin was still the language of higher and upper-class education, and people would learn other languages according to the necessities of their trade. So, for example, merchants involved in foreign trade would know a variety of languages depending on the direction of trade. For trade with continental Europe, French, Dutch, and Italian could be useful, while the more far-reaching trade of the East India Company, for example, made it useful for traders to learn at least a smattering of the languages of people traded with, as well as trading languages and jargons. The knowledge and use of languages other than English was reflected in the codeswitching patterns of different genres. There was a greater variety of languages switched into than in medieval times (see Raumolin-Brunberg, Chapter 45), reflecting the changes in society and types of texts, but Latin was still the most frequently occurring language. Particularly, scientific and religious texts show a high incidence of passages in languages other than English, especially when the intended audience was professional (Pahta and Nurmi 2006). This seems to indicate that there was still an expectation of Latin being known by the readers, even if the main body of the text was in English. Code-switching in these two domains can be seen as a bridge phenomenon in one of the processes that had a great influence on the development of Early Modern English, vernacularization (see Wright, Chapter 43). This was a progression that could be seen in many types of texts. While scientific and religious writing are often cited as examples, the same development could also be seen in e.g. administrative documents. English was now being used in registers and domains which had previously been performed in another language, most typically Latin and French. Go¨rlach (1999: 462) estimates that around 1500 legal texts were already mostly produced in English, although there were still remnants of Latin and French. In the realm of literature the rise of standard English is most evident, while some Latin and also dialects of English are still used. Scholarly texts are the area where Latin is still most frequently used. In Go¨rlach’s estimate, approximately half of scientific writing was in English at the beginning of the early modern period, the other half being mostly in Latin. The vernacularization process continued through the 16th and 17th centuries, and by the mid-17th century English was the primary language for scholarly texts in England. The language of religion also went through a gradual change. With the Reformation, Bible translations were ever more widely spread, and the language of liturgy changed from Latin to English. Because of the constant tension between Anglicans and Catholics and later also the Puritans and other groups, not to mention the rising Quakerism and other minor groups, there was also a constant need of discussion and writing on religion, and this was carried out in English, outlining the particularly English context in which these debates were carried out.
57
58
I Periods The expansion of English into new registers placed requirements on the language, and the influences can be seen on many levels of language. Not only does the lexicon constantly expand to accommodate the expression of new ideas, but new rhetorical styles had their influence on ideals of writing. The fact that education was still very much on the pattern of classical Latin meant that “the English style used in many formal text types was apparently praised according to how close it came to Latin models” (Go¨rlach 1999: 464). Even on the level of individual linguistic items the influence of education can be suspected. So, for example, epistemic uses of may and must spread first in the language of university-educated high-ranking men, which would suggest that the thought styles taught at Oxford and Cambridge included the use of epistemic modality (Nurmi 2003, 2009).
4 Resources for the study of Early Modern English Because of the proliferation of different types of text, we are also able to benefit from a larger variety of electronic corpora as sources for studying Early Modern English. On the one hand, we have multi-genre general-purpose corpora, and on the other, there are also more specialized corpora of a single genre or domain of writing. Of the first type, the Helsinki Corpus (Rissanen et al. 1991) covers the years 1500–1710 and consists of 500,000 words in 18 genres. The ARCHER corpus (Biber and Finegan 1990–93/2002/ 2007/2010) focuses more on the late modern period, but it starts from 1650, and has 11 genres. Also, other varieties are covered: ARCHER includes a corresponding selection of both British and American English, and the Helsinki Corpus of Older Scots (Meurman-Solin 1995) brings a possibility of contrastive studies. With the digitization of more and more materials, large commercial databases offer an ever increasing selection of the early printed sources in massive archives, such as Early English Books Online (Chadwyck-Healey 2003–2011) and the Literature Online (Chadwyck-Healey 1996–2011) database. The more specialized corpora focus on a single genre, topic domain, or publication type. A good example of the last is the Lampeter Corpus (Claridge et al. 1999). It contains tracts published between 1640 and 1740, and has six topic domains, which are represented for each decade of the timeperiod, numbering over a million words. The Zurich English Newspaper Corpus (ZEN) (Fries, Lehmann et al. 2004) covers early English newspapers between 1661–1791, giving access to 1.6 million words of whole newspaper issues with their varied content types. Newdigate Newsletters (Hines 1995) presents the precursor and competitor of the newspaper, written between 1674–1715 (750,000 words). The single-genre Parsed Corpus of Early English Correspondence (PCEEC) (Nevalainen et al. 2006) consists of personal letters written between 1410 and 1681, altogether 2.2 million words. There is also a short version containing a selection of the texts, the Corpus of Early English Correspondence Sampler (CEECS) (Kera¨nen et al. 1998), with 450,000 words. A Corpus of English Dialogues 1560–1760 (Kyto¨ and Culpeper 2006) focuses on speech-based texts, containing both authentic dialogue in trial proceedings and witness depositions, and constructed dialogue in drama and prose fiction, again reaching over a million words. There are also corpora focusing on a special domain of writing. The Corpus of Early English Medical Writing (Taavitsainen et al. 1995–) presents various text categories aimed at both expert and lay readership and covering multiple types of writing from
4 Periods: Early Modern English the purely academic to health guides for the general public. The Corpus of English Religious Prose (Kohnen et al. forthc.), which is being prepared, tackles the domain of religion and the various genres of writing that are connected with it. New corpus projects arise all the time, and the variety of these projects and the types of corpora they aim to build are a testimony to the multiplicity of material available for scholars of Early Modern English.
5 Changes in Early Modern English The two most striking changes taking place in the early modern period were the standardization, particularly of orthography, in published writing (with a gradual spread of similar spelling conventions to private texts as well) and the explosive growth of the lexicon. The early stages of descriptive and prescriptive writing on language were also seen during this period, even if the main developments only arrived in the 18th century. Many of the other linguistic developments of the age were continuations of long-term trends which had their origins in Middle and – in some cases – Old English. Standardization is most often viewed on the level of orthography, and certainly the changes there were remarkable during the two centuries in question, but also other levels of language can be argued to have developed some form of standard. Printing and the growing and developing civil service spread the particular type of writing of literate people in London and at the universities to a more varied readership than before, and provided a model to aim towards. Many linguistic features which become an established part of the new general dialect did not necessarily spread to the spoken regional forms of language, but, since our remaining sources are written, they tend to obscure the richness of local variation which must have existed all through the centuries in order to have survived to the present day (see Moessner, Chapter 44). The vocabulary of English was increasing as more types of texts were produced, and this led to hard word dictionaries being published. These often took the form of wordlists, which might contain words invented by the compiler of the lexicon, never seen outside these compilations, but they also presented many words which have since established their place in the English lexicon. Some dictionaries were aimed at translators, others specialized in a given field, such as legal or medical terminology or the language of thieves (see Considine, Chapter 66). The orthoepists discussed ways of improving the English spelling system and as an unintended side product gave us a clearer idea of how the language was pronounced. Early grammars were heavily based on the Latin model, and were often not very succesful in describing English in those terms. Because of the newly literate middle ranks of society, there was a welcoming readership for these works, although the age of the autodidact did not properly begin before the 18th century. How much influence any of these volumes had on the English actually used is an open question, but they give us an indication of the increasing interest in codifying, analyzing and teaching English. The English lexicon increased in size in several ways and for several reasons. New words were borrowed for new concepts, both scientific and cultural, from any number of languages. While Latin was the most influential source, the influx of new words from both European and world languages is notable. As the English became more familiar with the world, they introduced new words to describe the flora and fauna, the artifacts and merchandise, the peoples and cultures they encountered. At the same time, the
59
60
I Periods Renaissance ideal of expressing an idea in as many ways as possible contributed to the borrowing of Latin words in order to introduce variety. New words did not come solely from borrowing: also word formation through prefixes and suffixes, as well as compounding, was frequent. Many of the elements included in these processes were borrowed themselves, but there was eventually also mixing of native and borrowed elements. Many near synonyms were introduced, but only a selection of those has survived (see Lancashire, Chapter 40). On the level of morphology, the loss of nominal case endings that had been going on for a long time reached its culmination, with only the genitive -s remaining in the nominal system. The use of apostrophes to signify the genitive as distinguished from plural (or to distinguish genitive singular and plural) arose only gradually, and did not reach present usage before the 18th century. In the case of personal pronouns the most notable changes appeared with regard to second person. The singular pronoun thou became increasingly marked, and was used less and less except for highly specific contexts (intimacy, status difference, religious language). As thou disappeared, the corresponding verbal inflection disappeared as well. In the plural, the object form you replaced the old subject form ye. In the case of relative pronouns, subject pronoun who became established in human reference (see Busse, Chapter 46). Adjectives and adverbs showed more variation in the formation of comparative and superlative forms than Present-Day English, and the rules governing the use of inflections or the periphrastic forms were still in flux, leaving room for double forms (most happiest). Adverbs had variant forms without the suffix -ly, so that smooth/smoothly could be used interchangeably. The verbal system saw a rise of auxiliaries. Periphrastic do established itself in questions and negative statements. For a while, it seemed that do was also making inroads in affirmative statements, but this development was cut short (see Warner, Chapter 47). Verbal inflections followed a similar trend as nominal case endings, and the early modern period saw the loss of all but the third-person singular suffix, which changed from the earlier -th to -s. Since the loss of inflections made the subjunctive scarcer, modal auxiliaries took some of its functions. The meanings of modal auxiliaries shifted more towards the present model with the increasing frequency of epistemic meanings. The progressive be + -ing form started increasing, although the real development of this construction took place in the 18th and 19th centuries. As for word order, the long-term change towards a fixed pattern of subject-verbobject in declarative statements saw the last stages of formalization. Sentence-initial adverbs could still cause subject-verb inversion in the early modern period, but, apart from the greater liberties taken by verse, this pattern was notably less frequent by 1700. Many syntactic patterns typical of Latin could be seen in high styles of writing, whether legalese or ornate literature (see Seoane, Chapter 39). On the level of pronunciation, the Great Vowel Shift was perhaps the most notable development. The raising of long vowels took place over three centuries, and was a series of local developments (see Krug, Chapter 48). All parts of the shift did not run their course in all dialects, and there was variation in how individual words were affected. Local dialects continued as the main spoken form, but the beginnings of Received Pronunciation appeared in the cultural hub that was London (see Mugglestone, Volume 2, Chapter 121).
4 Periods: Early Modern English All in all, developments in Early Modern English levelled much of the earlier variation as the new standard language was formed. The place for standard was in official, published and formal kinds of writing, but private, unpublished, and informal language continued to show much more regional and stylistic variation. Being able to command the standard register was one of the requirements of inclusion in the elites of the country, but large parts of the population could lead successful lives without the requirements of shaping their language to this new pattern.
6 References 6.1 Online resources Biber, Douglas and Edward Finegan. 1990–93/2002/2007/2010. A Representative Corpus of Historical English Registers (ARCHER). Version 3.1 http://www.llc.manchester.ac.uk/research/ projects/archer/ Chadwyck-Healey. 2003–2011. Early English Books Online, 1475–1700 (EEBO). Ann Arbor: ProQuest. http://eebo.chadwyck.com/home Chadwyck-Healey. 1996–2011. Literature Online (LION). http://lion.chadwyck.com/ Claridge, Claudia, Josef Schmied, and Rainer Siemund. 1999. The Lampeter Corpus of Early Modern English Tracts. In: ICAME Collection of English Language Corpora (CD-ROM), 2nd edn., Knut Hofland, Anne Lindebjerg, and Jørn Thunestvedt (eds.), The HIT Centre, University of Bergen, Norway For manual see http://khnt.hit.uib.no/icame/manuals/LAMPETER/ LAMPHOME.HTM Fries, Udo, Hans Martin Lehmann et al. 2004. Zurich English Newspaper Corpus (ZEN). Version 1.0. Zu¨rich: University of Zu¨rich. http://es-zen.unizh.ch (see also: http://www.helsinki.fi/ varieng/CoRD/corpora/ZEN/index.html) Hines, Philip, Jr. 1995. Newdigate Newsletters. In: ICAME Collection of English Language Corpora (CD-ROM), 2nd edn., Knut Hofland, Anne Lindebjerg, and Jørn Thunestvedt (eds.), The HIT Centre, University of Bergen, Norway. For manual see http://icame.uib.no/newdigateeks.html Kera¨nen, Jukka, Minna Nevala, Terttu Nevalainen, Arja Nurmi, Minna Palander-Collin and Helena Raumolin-Brunberg. 1998. Corpus of Early English Correspondence Sampler (CEECS). In: ICAME Collection of English Language Corpora (CD-ROM), 2nd edn., Knut Hofland, Anne Lindebjerg, and Jørn Thunestvedt (eds.), The HIT Centre, University of Bergen, Norway. For manual see http://khnt.hit.uib.no/icame/manuals/ceecs/INDEX.HTM (see also: http://www. helsinki.fi/varieng/domains/CEEC.html) Kohnen, Thomas, Sandra Boggel, Tanja Ru¨tten, Dorothee Groeger, Ingvilt Marcoe, and Kirsten Gather. forthc. Corpus of English Religious Prose (COERP). http://www.helsinki.fi/ varieng/CoRD/corpora/COERP/index.html Kyto¨, Merja and Jonathan Culpeper. 2006. A Corpus of English Dialogues 1560–1760. With the assistance of Terry Walker and Dawn Archer. Uppsala University and Lancaster University. http://www.engelska.uu.se/Research/English_Language/Research_Areas/Electronic_Resource_ Projects/A_Corpus_of_English_Dialogues/ Meurman-Solin, Anneli. 1995. Helsinki Corpus of Older Scots. Department of English, University of Helsinki. http://www.helsinki.fi/varieng/CoRD/corpora/HCOS/index.html Nevalainen, Terttu, Helena Raumolin-Brunberg, Jukka Kera¨nen, Minna Nevala, Arja Nurmi, and Minna Palander-Collin. 2006. Parsed Corpus of Early English Correspondence (PCEEC). Annotated by Ann Taylor, Arja Nurmi, Anthony Warner, Susan Pintzuk, and Terttu Nevalainen. Helsinki: University of Helsinki and York: University of York. Distributed through the Oxford Text Archive. Rissanen, Matti, Merja Kyto¨, Leena Kahlas-Tarkka, Matti Kilpio¨, Saara Nevanlinna, Irma Taavitsainen, Terttu Nevalainen, and Helena Raumolin-Brunberg. 1991. The Helsinki Corpus of English Texts. In: ICAME Collection of English Language Corpora (CD-ROM), 2nd edn.,
61
62
I Periods Knut Hofland, Anne Lindebjerg, and Jørn Thunestvedt (eds.), The HIT Centre, University of Bergen, Norway. For manual, see http://khnt.hit.uib.no/icame/manuals/hc/index.htm Taavitsainen, Irma, Pa¨ivi Pahta, Martti Ma¨kinen, Turo Hiltunen, Ville Marttila, Maura Ratia, Carla Suhr, and Jukka Tyrkko¨. 1995–. Corpus of Early English Medical Writing (CEEM). University of Helsinki. http://www.helsinki.fi/varieng/CoRD/corpora/CEEM/index.html
6.2 Printed resources Barber, Charles. 1997. Early Modern English. 2nd edn. Edinburgh: Edinburgh University Press. Barry, Jonathan. 1995. Literacy and literature in popular culture: Reading and writing in historical perspective. In: Tim Harris (ed.), Popular Culture in England, c. 1500–1850, 69–94. London: Macmillan. Briggs, Asa. 1994. A Social History of England: From the Ice Age to the Channel Tunnel. 2nd edn. London: Weidenfeld and Nicolson. Brink, Jean R. 2010. Literacy and education. In: Hattaway (ed.), Vol. 1, 27–37. Coleman, David and John Salt. 1992. The British Population: Patterns, Trends, and Processes. Oxford: Oxford University Press. Cressy, David. 1975. Education in Tudor and Stuart England. London: Arnold. Cressy, David. 1980. Literacy and the Social Order: Reading and Writing in Tudor and Stuart England. Cambridge: Cambridge University Press. Go¨rlach, Manfred. 1991. Introduction to Early Modern English. Cambridge: Cambridge University Press. Go¨rlach, Manfred. 1999. Regional and social variation. In: Lass (ed.), 459–538. Hattaway, Michael (ed.). 2010. A New Companion to English Renaissance Literature and Culture. Vols. 1–2. Oxford: Wiley-Blackwell. Kishlansky, Mark. 1996. A Monarchy Transformed: Britain 1603–1714. London: Penguin. Lass, Roger. 1999. Introduction. In: Lass (ed.), 1–12. Lass, Roger (ed.). 1999. The Cambridge History of the English Language. Vol. III. 1476–1776. Cambridge: Cambridge University Press. Milroy, James. 1992. Linguistic Variation and Change. Oxford: Blackwell. Nevalainen, Terttu. 2006. Introduction to Early Modern English. Edinburgh: Edinburgh University Press. Nurmi, Arja. 2003. The role of gender in the use of MUST in Early Modern English. In: Sylviane Granger and Stephanie Petch-Tyson (eds.), Extending the Scope of Corpus-based Research: New Applications, New Challenges, 111–120. Amsterdam/Atlanta: Rodopi. Nurmi, Arja. 2009. May: The social history of an auxiliary. In: Andreas H. Jucker, Daniel Schreier, and Marianne Hundt (eds.), Corpora: Pragmatics and Discourse. Papers from the 29th International Conference on English Language Research on Computerized Corpora (ICAME 29). Ascona, Switzerland, 14–18 May 2008, 321–342. Amsterdam/Atlanta: Rodopi. O’Callaghan, Michelle. 2010. Publication: Print and manuscript. In: Hattaway (ed.), Vol. 1, 160– 176. Pahta, Pa¨ivi and Arja Nurmi. 2006. Code-switching in the Helsinki Corpus: A thousand years of multilingual practices. In: Nikolaus Ritt, Herbert Schendl, Christiane Dalton-Puffer, and Dieter Kastovsky (eds.), Medieval English and its Heritage: Structure, Meaning and Mechanisms of Change, 203–220. Frankfurt: Peter Lang. Palander-Collin, Minna. 2009. Patterns of interaction: Self-mention and addressee inclusion in the letters of Nathaniel Bacon and his correspondents. In: Arja Nurmi, Minna Nevala, and Minna Palander-Collin (eds.), The Language of Daily Life in England (1400–1800), 53–74. Amsterdam/Philadelphia: John Benjamins. Rappaport, Steve. 1989. Worlds within Worlds: Structures of Life in Sixteenth-century London. Cambridge: Cambridge University Press.
5 Periods: Late Modern English
63
Raumolin-Brunberg, Helena. 1998. Social factors and pronominal change in the seventeenth century: The Civil-War effect? In: Jacek Fisiak and Marcin Krygier (eds.), Advances in English Historical Linguistics (1996), 361–388. Berlin: Mouton de Gruyter. Youings, Joyce. 1984. Sixteenth-century England. London: Penguin.
Arja Nurmi, Helsinki (Finland)
5 Periods: Late Modern English 1 2 3 4 5 6 7 8 9
Introduction: definitions of Late Modern English The growth of Late Modern English studies External history Syntax and morphology Phonology Lexis Normative grammarians Summary References
Abstract This chapter provides an overview of research into the history of English in the Late Modern period. It begins with an account of how this period came to be defined as a distinct period lasting roughly from 1700–1900 and goes on to discuss the reasons why this period has, until relatively recently, received less scholarly attention than earlier ones. An overview of the external history of the period follows, concentrating on factors such as urbanization, industrialization, and the growth of transport and communications technology, all of which contributed to the social and geographical mobility which characterize this period. The remainder of the chapter provides brief accounts of research into the syntax and morphology, phonology, and lexis of this period. The chapter concludes with a short account of recent research which re-evaluates the normative grammarians of this period and some suggestions for future directions for the study of Late Modern English.
1 Introduction: definitions of Late Modern English The phrase “Late Modern English” seems to have been first used by Poutsma in the title of his Grammar of Late Modern English (first published in 1914), but he was referring here not to a historical period so much as to what was, to him, contemporary English, “the English Language as it presents itself in the printed documents of the last few generations” (Poutsma 1928: viii). As such, his study was synchronic rather than diachronic, dealing with the language as it was in his time and in the very recent past. The tripartite division of the history of English into Old, Middle and Modern English can probably be attributed to Sweet, who proposed this division in a lecture to the Philological Society (1873–1874). Sweet saw “Modern English” as a unity stretching from the 16th century to his own time and characterized it as the period of lost inflections. Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 63–78
64
I Periods The recognition of Late (or Later) Modern English as a specific period in the history of English appears to have followed much later. Charles Barber uses the term “Later Modern English” to distinguish this period from Early Modern English: I am taking eModE to be the English language between 1500 and 1700. All such divisions are arbitrary, for linguistic change is continuous; but there are a number of features in the language of that period which mark it off fairly clearly from Middle English (ME) and Later Modern English (LModE) (Barber, 1976: 1).
Barber provides no end-point for the Later Modern period, but a consensus has since emerged that the Late Modern period lasts roughly from 1700 to 1900, though Beal (2004) extends the end-point to 1945. (See Curzan, Volume 2, Periodization.)
2 The growth of Late Modern English studies Charles Jones refers to the 18th and 19th centuries as the “Cinderellas of English historical linguistic study” (Jones 1989: 279). This situation lasted until the final decade of the 20th century, when Bailey (1996) and Go¨rlach (1999) both published monographs on 19th-century English. Until this point, studies of 18th- and/or 19th-century English tended to concentrate on specific areas. The agenda for research on this period had been set by Leonard’s (1929) volume on “correctness” in 18th-century English, so that general histories of English tended to describe this period as characterized by the appearance of prescriptive grammars and authoritative dictionaries. These earlier works express the view that there were no linguistic changes worthy of investigation after 1700. A typical statement is that made by Bloomfield and Newmark, who assert that “after the period of the Great Vowel Shift was over, the changes that were to take place in English phonology were few indeed” (Bloomfield and Newmark 1963: 293), and, even as late as 1992, Freeborn stated that “the linguistic changes that have taken place from the eighteenth century to the present day have been relatively few” (Freeborn 1992: 180). That any language could be spoken for 200 years without any change taking place is highly unlikely, but scholars were referring here to the kind of major, structural changes, such as the Great Vowel Shift or the introduction of do-support, which dominate much discussion of earlier periods. Some scholars, such as Strang, suggest that the nature of change in the Late Modern period was different from that in Middle and Early Modern English. Some short histories of English give the impression that changes in pronunciation stopped dead in the 18c, a development which would be inexplicable for a language in everyday use. It is true that the sweeping systematic changes we can detect in earlier periods are missing, but the amount of change is no less. Rather, its location has changed: in the past two hundred years changes in pronunciation are predominantly due, not, as in the past, to evolution of the system, but to what, in a very broad sense, we may call the interplay of different varieties, and to the complex analogical relationship between different parts of the language (Strang 1970: 78–79).
Although Strang acknowledges that changes have taken place since 1700, her suggestion that “sweeping systematic changes” were completed by then and that the later changes were caused by “the interplay of different varieties” would appear to flout
5 Periods: Late Modern English the “uniformitarian principle” that the mechanisms governing linguistic variation and change operated in past times and societies as they do today. Research in the fields of socio-historical linguistics and historical sociolinguistics, such as Romaine (1982) and Nevalainen and Raumolin-Brunberg (2003), has demonstrated that “the interplay of different varieties” was as much a factor in linguistic change in the Early Modern period as Strang claims it to be for “the past two hundred years”. The perceived difference between linguistic changes of earlier and later periods is more likely to be a result of the scholar’s perspective: changes appear simpler and more abstract as chronological distance increases, and as the amount of detailed evidence for variation decreases. Jones suggests as much when he writes: There has always been a suggestion … especially among those scholars writing in the first half of the twentieth century, that phonological and syntactic change is only properly observable at a great distance and that somehow the eighteenth, and especially the nineteenth centuries, are “too close” chronologically for any meaningful observations concerning language change to be made (Jones, 1989: 279).
Perhaps one reason for the upsurge of interest in Late Modern English in the 1990s is that, as the new millennium hove into sight, scholars felt sufficiently distanced from the 18th and 19th centuries to be able to observe the linguistic changes that occurred in this period. However, chronological distance is not the only reason for the growth of studies in Late Modern English in the late 20th and early 21st centuries. The development of corpus studies and the availability of a wide range of texts and text types in electronic form have enabled scholars both to identify overall patterns of change and to interrogate individual texts in ways which were simply not possible prior to the digital revolution. Historical linguists can now search for syntactic patterns or lexical collocations in corpora consisting of millions of words, or they can access individual texts for close study. David Denison’s comment on the nature of syntactic change in Late Modern English comes from this perspective: Since relatively few categorical losses or innovations have occurred in the last two centuries, syntactic change has more often been statistical in nature, with a given construction occurring throughout the period and either becoming more or less common generally or in particular registers. The overall, rather elusive effect can seem more a matter of stylistic than syntactic change (Denison 1998: 93).
The area of linguistics in which constructions becoming “more or less common” is of primary interest is, of course, sociolinguistics, where quantitative methods have been used to investigate variation and change in language since the 1960s. The application of models taken from sociolinguistics has allowed scholars to make sense of the variability of data from the Late Modern period. The notion of “stylistic change” has also been given more prominence in sociolinguistics than in formal linguistics, and the application of sociolinguistic models, together with the availability of texts, has led to the investigation of the “styles” of social networks and of individuals in the Late Modern period. So, we could say that a confluence of circumstances has led to the increasing prominence of Late Modern English studies in the late 20th and early 21st centuries. There
65
66
I Periods are now two monographs entirely devoted to this period, Beal (2004) and Tieken-Boon van Ostade (2009); a conference on Late Modern English initiated by Charles Jones in 2001 has now had four meetings; and the proceedings of the first three of these conferences have been published as Dossena and Jones (eds.) (2003), Bueno Alonso et al. (eds.) (2007), and Tieken-Boon van Ostade and van der Wurff (eds.) (2009). The contents of these volumes provide a good indication of the agenda that is developing for those working in the field of Late Modern English. Dossena and Jones divide their volume into three sections entitled “The Late Modern English Grammatical Tradition”, “The Syntax of Late Modern English”, and “Language and Context in the Late Modern Period”. The third section is fairly eclectic, dealing with the language of individuals, and with pragmatics and sociolinguistics. Bueno Alonso et al. present individual contributions rather than grouping them into sections, but of the fifteen papers, seven are devoted to syntax or morphology, three to pragmatics, and five to areas such as variation and change and the development of genres. Tieken and van der Wurff divide their volume into four sections entitled “Prescriptive and Normative Concerns”, “Late Modern Work on the English Language”, “Studies in Grammar and Lexis”, and “Studies on Letters”. This indicates that the main areas of research in Late Modern English are: morpho-syntactic change; the normative tradition; historical pragmatics; and what we might broadly refer to as historical sociolinguistics. This is not to say that other areas have been totally neglected: for instance, two monographs have been published (Beal 1999 and Jones 2006) which are entirely devoted to the phonology and phonetics of English in this period. As the quotes from Strang (1970) and Denison (1998) above might have predicted, the study of Late Modern English has more recently been informed by the methods and frameworks of sociolinguistics and corpus linguistics (see, for instance, Mair 2002; Rohdenburg 2007; and De Smet 2007), so that research is often difficult to categorize in conventional linguistic terms. In the sections that follow, I shall provide a brief account of the external history of English-speaking communities in the Late Modern Period, then go on to summarize the main developments in morpho-syntax, phonology, and lexis, before discussing the ways in which scholars of Late Modern English have interrogated the normative texts of this period to arrive at a more nuanced view of the “Age of Authority” than had hitherto been put forward.
3 External history We saw in Section 1 that the Late Modern period is usually described as beginning around 1700. In his introduction to Volume III of the Cambridge History of the English Language covering the period 1476–1776, Lass gives a dramatic summary of the events leading up to this point in history: “By the eighteenth century, the nation had been through a religious reformation, a regicide, a commonwealth, the flight of the hereditary monarch, and the accession of a foreign king who signed away much of his power” (Lass 1999: 3). Lass also agrees with the historian Roy Porter (2000) in viewing the publication of Newton’s Principia (1687) as marking the beginning of the English Enlightenment, an age of “reason” and science. The most obvious linguistic consequence of scientific progress in the Late Modern period was to stimulate lexical innovation as new inventions, processes, and whole disciplines required names. Newton wrote his Principia in Latin, but the Royal Society was to foster a style of scientific writing in English as writing in Latin declined.
5 Periods: Late Modern English The scientific discoveries of the late 17th and early 18th centuries led to the technological innovations which drove the Industrial Revolution of the late 18th and 19th centuries. Britain became an industrial and urban nation as workers moved from the countryside to the newly-expanding towns and cities. The historian Michael Rose writes: At the beginning of the [nineteenth] century there were only fifteen towns in England and Wales with more than 20,000 inhabitants; by 1851, there were sixty-three, and one half of the population could be described as town dwellers compared to about one third in the late eighteenth century. By 1900, almost 80 per cent of the population lived in urban districts with populations of 10,000 or more. In the space of a hundred years, Britain had transformed into an urban society (Rose 1985: 277).
The main linguistic consequence of this urbanization and movement of populations was dialect contact, leading to levelling and the formation of new, urban dialects. Dialect contact was also facilitated by advances in transport and communications in this period. In the course of the 18th century, the Turnpike Trusts funded a substantial number of new roads, cutting the length of a journey by carriage from York to London from three days to one, and opening up the possibility of travel for leisure as well as business, at least for the reasonably well-to-do. The development of the railway network in the 19th century made affordable leisure travel possible for the lower and middle classes. The introduction of the Penny Post in 1840 and the electric telegraph in 1837 increased the possibilities for written communication, and bequeathed a rich legacy of data to scholars of Late Modern English, who have been able to create corpora of letters from this period, such as the Corpus of late 18c Prose (Denison et al. 2002) and the 18th century extension to the Corpus of Early English Correspondence (CEECE) (Nevalainen et al. 1998–2006). The invention of the phonograph in 1877 made it possible for speakers of English to hear the disembodied voices of speakers from distant places. All these developments had the effect of increasing dialect contact between speakers (and writers) from different parts of Britain (and beyond). Dialect contact of a different kind was occasioned by the introduction of compulsory, free elementary schooling in 1870. One consequence of this was that every child would come into contact with Standard English because every village would have at least one schoolteacher who, along with the clergyman, would act as models of “correct” usage. Scholarly interest in philology led to the formation of the English Dialect Society in 1873, the objective of which was to collect material for a comprehensive dictionary of English dialects. This objective was achieved with the publication of Wright’s English Dialect Dictionary (1898–1905), but in the interim many glossaries of individual dialects were published under the auspices of the English Dialect Society. The authors of these glossaries all express a sense of urgency: they have collected their material “just in time”, and all blame universal education and the railways for the imminent demise of dialects. An example of such a statement is the following, from Morris’s Yorkshire Folk-Talk: Railways and certificated schoolmasters, despite their advantages, are making sad havoc of much that is interesting and worth preserving in the mother tongue of the people. This is to be regretted. It is with the object of collecting any such relics of the past, which would otherwise be doomed to oblivion, that I make the following appeal to my brother Yorkshiremen (Morris 1892: v).
67
68
I Periods What few of these 19th-century antiquarians and philologists recognized was that dialect contact can create new varieties as well as destroying old ones. In the towns and cities of the Industrial Revolution, new, urban dialects were being forged in the crucible of contact between speakers of different regional and national varieties of English. In some places the impact of Irish immigration during and immediately after the Great Famine was considerable: in Liverpool and in the 19th-century “new town” of Middlesbrough, approximately one-fifth of the population recorded in censuses of this period was Irish-born. Elsewhere, the influx of population was mainly from the rural hinterland, but in places such as Newcastle and the West Riding of Yorkshire, dialect writing and performance testifies to a growing consciousness of, and pride in, new urban identities and varieties of speech (Beal 2000, 2011). These urban varieties have since become the main object of dialect study in the late 20th and early 21st centuries, and are now in their turn seen as potentially threatened by levelling and supralocalization. The Late Modern period also saw the beginning of the “great divide” between British and American English. Although the first English-speaking colonies in what is now the USA were founded in the early 17th century, the development of American English as a national variety with its own prescribed norms was precipitated by the American Revolution (1775–1783). In 1789, Webster asserted that “customs, habit and language, as well as government, should be national. America should have her own, distinct from all the world” (Webster 1789: 179). His American Dictionary of the English Language (Webster 1828) provided norms for spelling which were deliberately differentiated from those of British English, and legitimized “Americanisms” by including them in the dictionary. Once America became independent, Americans loyal to the crown moved to Canada and British colonial expansion diverted to Australia in 1788 and, in the 19th century, to South Africa and New Zealand. The development of distinct national norms of English in these countries was perhaps more a phenomenon of the 20th century, but the history of these Englishes belongs entirely in the Late Modern period (see Volume 2, Section XV). Thus, the Late Modern period is one in which English ceases to be predominantly the “property” of speakers in Britain, and any history of English in this period needs to take account of this. Another linguistic consequence of the British Empire was the introduction into English of loan words from a wide variety of languages, as flora, fauna, topographic features, and customs hitherto unknown to speakers of English required names. This, along with the scientific discoveries and inventions referred to above, accounts for the dramatic increase in lexical innovation during the 19th century. Within Britain, the commercial opportunities of the Industrial Revolution and the expansion of educational provision led in the course of the Late Modern period to the emergence of an ambitious and influential middle class, who were now able to gain the wealth, power, and influence which had formerly been the preserve of the landed gentry. Indeed, the most prominent industrialists and entrepreneurs could be ennobled to the peerage, build their own “great houses” in the country, socialize with the gentry, and send their sons to the public schools which became the cradle of Received Pronunciation. Social mobility led to linguistic insecurity amongst the upwardly-mobile and so created a market for the normative texts for which this era is famous. As has already been noted, Leonard’s (1929) work on the grammarians of the 18th century was so influential that, for much of the 20th century, the received view of these grammars was that they imposed arbitrary rules to suppress variation in English. In recent years, the wider availability of these texts facilitated by Eighteenth-Century
5 Periods: Late Modern English Collections Online (Gage Cengage Learning 2009) and other digitization projects, has allowed scholars to develop a more nuanced account of these grammars, and to compare precept with practice. Nevertheless, it cannot be denied that the Late Modern period marks the final stages of the standardization process for British English and that the prescriptive attitudes of this era are still with us today (Beal 2008, 2009).
4 Syntax and morphology As Denison noted (see Section 2), the Late Modern English period saw very little in the way of categorical change in syntax or morphology. Some constructions and patterns which had been used variably in Early Modern English became regulated in this period, so that variants formerly found in Standard English texts would be restricted to nonstandard usage and/or “marked” in some way. Many scholars have blamed the prescriptive grammarians of the 18th century for this regulation, but recent research involving the comparison of precept and practice in this period suggests that they were simply codifying what was already best practice. I shall therefore discuss these changes in Section 7, and concentrate here on the more categorical changes. As far as morphology is concerned, there was very little change within Standard English in the Late Modern period. We saw in Section 1 that Sweet defined the Modern English period as a whole as the period of lost inflections. However, the only inflection to be lost after 1700 is the second person singular -st ending. This in turn depends on the loss, from Standard English, of the distinction between second person singular thou, thee, thy, thine and the formerly plural ye, you, your, yours. The singular forms had become marked in the Early Modern period, and by 1700 “survived only in dialects, among Quakers, in literary styles, as a device of heightening […] and in its present religious function” (Strang 1970: 140). Some 18th-century authors, notably Sheridan and Richardson, put thou forms into the mouths of upper-class males, but this usage was not universally accepted. Although 18th-century grammarians often include thee, thou, thy in their paradigms of pronouns in order to demonstrate the singular-plural distinction, Greenwood tells his readers that “it is ungentile [sic.] and rude to say, Thou dost so and so” (Greenwood 1711: 110). Later in the century, Lowth states that “Thou in the Polite, and even in the Familiar style, is disused, and the plural You is employed instead of it” (Lowth 1762: 48). In this case, the statements of the grammarians accord with the evidence from 18th-century texts: by the middle of the century, use of thou had declined from “ungentile” to “disused”, at least in Standard English. (See further Busse, Chapter 46.) The only categorical innovation in Late Modern English syntax concerns what has variously been called the “be + -ing” construction, the “progressive” and the “expanded form”, as in She is reading a book. Although this construction had been used before 1700, its use in Early Modern English was optional in contexts where today it would be required. In Hamlet (II.ii.190), Polonius asks Hamlet “What do you read my Lord?” Today this would be interpreted as an inquiry into the prince’s reading habits, but Polonius was referring to the book that Hamlet had in his hands at the time. Today, the required construction in this pragmatic context would be “what are you reading, my Lord?” There was a remarkable increase in usage of this construction throughout the Late Modern period, both in terms of the sheer numbers of such constructions, and the types of clause in which it can occur, such that it becomes fully grammaticalized in the course of this period. From the second half of the 18th century, there is a rise
69
70
I Periods in its use with stative verbs such as love, wish, etc., with verbs denoting “instant” actions such as explode, fall etc., and with nominal and adjectival complements (e.g. You’re being a fool/foolish). The extension to the passive is likewise an innovation of the Late Modern period. Until the late 18th century, passive voice and progressive aspect could not both be marked in a clause, so that, if somebody wished to say that a house was under construction but not yet finished, the most acceptable way of expressing this would be “the house is building”. The passive was understood and the sentence unambiguous because houses can’t build themselves. The first examples of the passive with be + -ing are found in late 18th-century letters (1): (1)
I have received the speech and address of the House of Lords; probably, that of the House of Commons was being debated when the post went out. (Letters “to” Ser. Lett. 1st Earl Malmesbury; Denison 1998: 152).
Letters are, of course, the most informal genre of writing, and innovations are likely to be recorded earlier here than in more formal styles. It took some time for this construction to become accepted in all styles: even in the early 20th century Curme and Kurath seem to begrudge it: From 1825 on […] the form with being + perfect participle began to lead all others in this competition, so that in spite of considerable opposition the clumsy is being built became more common than is building in the usual passive meaning, i.e. where it was desired to represent a person or thing as affected by an agent working under resistance vigorously and consciously to a definite end: “The house is being built”, “My auto is being repaired” (Curme and Kurath 1931: 444).
The extension of the construction to longer verb phrases involving perfective and modal verbs came later. They appear in the late 19th century, but are roundly condemned by grammarians. By the early 20th century, such constructions are found in literary texts: they were rare then as now, simply because the pragmatic circumstances in which they might be used are rare. An early example is from Galsworthy (2): (2)
She doesn’t trust us: I shall always be being pushed away from him by her. (1915 Freelands; Denison 1998: 158).
A much fuller account of this and other syntactic changes of the Late Modern period can be found in Aarts, Lo´pez-Cuoso, and Me´ndez-Naya, Chapter 54. I have singled out the “be + -ing” construction here because it is one of very few areas in which the grammar of English has changed between 1700 and the present day to the extent that an earlier usage would seem ungrammatical to today’s speakers and, conversely, it is difficult for us to comprehend what 18th- and 19th-century grammarians found so “unnatural” in a sentence such as the house is being built.
5 Phonology The phonology of Late Modern English has, until very recently, had much less scholarly attention paid to it than that of earlier periods. This is probably because, as MacMahon
5 Periods: Late Modern English suggests, “superficially, the period under consideration might appear to contain little of phonetic and phonological interest, compared with, for example, earlier changes such as the transition from Old to Middle English, and the Great Vowel Shift” (MacMahon 1998: 373). However, from the sociohistorical point of view, it is a very rich period. The 18th century saw the beginning of the elocution movement and the publication of pronouncing dictionaries intended as guides to the “correct” pronunciation of English, i.e. that of educated, well-bred Londoners, whilst the 19th century witnessed the rise of Received Pronunciation (RP) (see Mugglestone, Volume 2, Chapter 121). Although the pronouncing dictionaries written by elocutionists such as Walker (1791) and Sheridan (1780) were undoubtedly normative, their detailed descriptions of sounds and transcriptions of every word in their dictionaries provide a wealth of evidence for the prestigious pronunciation of the period. A more detailed account can be found in Beal (1999) and Beal (2004: 125–167); see, also, Jones, Chapter 52: here, I shall briefly describe the main changes in the pronunciation of “received” English between 1700 and 1900. As far as consonants are concerned, the most striking development, at least as far as the pronunciation of English in England is concerned, is the loss of rhoticity. This is one of the most salient perceived differences between British and American English today (even though many varieties of British English are rhotic and several varieties of US English are non-rhotic). The detailed evidence from 18th- and 19th-century sources reveals that this change came “from below” and that it took a century for it to become accepted in the usage of RP speakers. Walker is often cited as the earliest source of evidence for the loss of rhoticity: “In England, and particularly in London, the r in lard, bard, card, regard, is pronounced so much in the throat, as to be little more than the middle or Italian a lengthened into baa, baad, caad, regaad” (Walker 1791: 50). But a careful reading reveals that he considered this to be a marker of lower-class London usage. Moreover, Walker describes the Irish pronunciation of /r/ as too harsh, but says that the pronunciation at the beginning of a word should be more “forcible” than at the end, so that “Rome, river, rage, may have the r as forcible as in Ireland, but bar, bard, card, hard &c. must have it nearly as soft as in London” (Walker 1791: 50). This suggests Walker was recommending a weakened /r/ in final and preconsonantal positions rather than total loss of rhoticity. As Mugglestone (1995: 98– 103) demonstrates, “dropping” of /r/ continued to be overtly stimatized until the late 19th century. Today, of course, it is the rhotic accents of England that are stigmatized as “rustic”. The other main consonantal changes in Late Modern English are not so much changes in the system, or even the distribution of phonemes, as the regulation of variants. Two of the greatest shibboleths of non-standard pronunciation in the 20th and 21st centuries are popularly known as “dropping” of and . In the latter case, the term “dropping” is not at all accurate, since the stigmatized variant is /n/ as opposed to /ŋ/ in, e.g. hunting, shooting and fishing. In both cases, the stigmatized variants had been attested at least from the Early Modern period, but are not labelled as “vulgar” or “incorrect” before the 18th century. Whilst “h-dropping” became the greatest social shibboleth of the 19th century, the alveolar pronunciation of has a more complex history. This pronunciation was a marker both of lower-class and upper-class usage throughout the Late Modern period and it can still be heard in the speech of very elderly, very conservative RP speakers. The stigmatization of alveolar thus
71
72
I Periods provides an intriguing example of a linguistic change of the Late Modern period in which middle-class speakers were in the vanguard, as has so often been seen to be the case in sociolinguistic studies of the 20th century. Such vowel changes as occurred in the Late Modern period largely involve the continuation of processes begun in the 16th and 17th centuries. In all cases, the earlier variants are still found in English regional accents, with many of the innovations still confined to RP and southern English varieties. The innovation known as the STRUT– ¯ and u˘ were unrounded to /ʌ/ in southern FOOT split, whereby some reflexes of ME o accents of (English) English, can be traced back to the 17th century. However, the lack of this split became stigmatized as “provincial” in the 18th century (and remains so to this day). Walker writes: If the short sound of the letter u in trunk, sunk etc., differ from the sound of this letter in the northern parts of England, where they sound it like the u in bull […] it necessarily follows that every word where this letter occurs must by these provincials be mispronounced (Walker 1791: xiii).
The other highly salient marker of the “north-south divide” in present-day accents of England, the pronunciation of the vowel in bath, laugh, grass, etc., has a more complex history. The lengthening and backing of this vowel seems, like the loss of rhoticity, to have been a change “from below”. Evidence of lengthening of /a/ in certain environments can be found in late 17th-century sources, but, throughout the 18th and 19th centuries, the pronunciation with /ɑː/ was not universally accepted. Walker tells us that, although “Italian a” had previously been heard in words such as glass, fast: “this pronunciation seems to have been for some years advancing to the short sound of this letter, as heard in hand, land, grand etc. and pronouncing the a in after, answer, basket, plant, mast, etc. as long as in half, calf etc. borders very closely on vulgarity” (Walker 1791: 10). This change seems to have begun as a lengthened /æː/ in the 17th century, and not to have become stigmatized until the lengthened vowel was retracted to [ɑː] The latter pronunciation is described as “drawling” throughout the 19th century, and there is evidence of a pronunciation with [æ] or even [ɛ] by young ladies wanting to avoid the “vulgar” [ɑː]. Those who wished their pronunciation to be beyond reproach had to avoid both the “drawling” [ɑː] and the “mincing” [æ] at least until the beginning of the 20th century, when Daniel Jones’s use of cardinal [ɑ] seems to have established this as the RP pronunciation. The other vowel changes to be considered here could be regarded as the tail-end of the Great Vowel Shift. In words such as face and goat, the ME vowels had been raised to /eː/ and /oː/ respectively, and these are the pronunciations recommended by Walker, who describes them as the “long, slender” sound of and the “long, open” sound of respectively, both of which are unequivocally monophthongal. However, the first evidence for diphthongal pronunciations of both these vowels comes very soon after Walker’s first edition: MacMahon (1998: 459) points out that the first evidence for diphthongization of /oː/ comes from the Scottish orthoepist William Smith in 1795 and it is generally accepted that the first attestation of a diphthongal pronunciation of /eː/ comes from Batchelor (1809). (In both cases, the diphthongal pronunciations are widely accepted in the 19th century, and are still found in RP and many other accents of present-day English.)
5 Periods: Late Modern English
73
6 Lexis The picture that we have of lexical innovation in the Late Modern period has largely been determined by the policies and practices of the Oxford English Dictionary (OED) (Simpson [ed.] 2000–). Until very recently, any research of a statistical nature into lexical innovation in this period (and earlier ones) was based on the Chronological English Dictionary (CED) (Finkenstaedt et al. 1970), which in turn took its data from the Shorter Oxford English Dictionary [SOED] [Onions 1964]). The CED has proved immensely useful to historians of English, as it reorganizes entries in the SOED by year of first citation and provides a numerical code to indicate the etymology of each word. For each year, the total number of first citations is given at the end of the table. Figure 5.1 is based on the total number of first citations for each decade of the Late Modern Period. 3000 2500 2000 1500 1000 500
0
0 –1
90
0
89
–1 18
91
0
88
–1 18
81
0
87
–1
61
18
18
71
0 –1
86
0
85
–1
41
18
18
51
0 –1
84
0
83
–1 18
31
0
82
–1
11
18
18
21
0 –1
81
0
80
–1
91
18
17
01
0 –1
79
0
78
–1
71
17
17
81
0 –1
77
0
76
–1 17
61
0
75
–1
41
17
17
51
0 –1
74
0
73
–1
21
17
17
31
0 –1
72
0
71
–1 17
11
0
70
–1 17
01
0
69 16
91
–1
68 81
–1 16
71
16
16
61
–1
67
0
0
Figure 5.1: Numbers of first citations in SOED by decade 1661–1900
What is immediately apparent from Figure 5.1 is that lexical innovation is at a low rate through the late 17th and 18th centuries, and then rises to a peak in the mid-19th century before a further sharp decline towards the end of the century. However, the authors of the CED acknowledge that “the vocabulary of the twentieth century is less systematically represented than that of the preceding periods” (Finkenstaedt et al. 1970: xi). This almost certainly accounts for the steep drop shown in Figure 5.1 from the end of the 19th century onwards: since the SOED is based on the first edition of the OED, it reflects the coverage of the latter, work on which began in earnest in 1879, but which was not published in its entirety until 1928. It is hardly surprising that innovations of the late 19th and early 20th centuries were not recorded, as the earlier fascicles, covering earlier letters of the alphabet, would have been completed by then. The apparent “trough” in the 18th century can also, to some extent, be accounted for by the practices of the OED, as the compilers of the first edition excerpted texts from the 18th century much less exhaustively than those from earlier periods or from the 19th century. This is now being remedied by the compilers of the third edition of the OED, so that research based on the OED online will give a more accurate picture
74
I Periods of lexical innovation in this period. Nevertheless, it still seems to be the case that there was less innovation in the 18th century than in the early 17th or mid-19th centuries. It is tempting to attribute this decline in innovation to the conservatism of the age. Celebrated authors such as Swift and Addison ridiculed the pretentious innovations of the age in the periodicals of the time such as the Tatler and the Spectator, but the same kind of discourse can be found in today’s newspapers whenever a dictionary releases a list of new words. If there was no innovation, they would have nothing to satirize. It is more likely that the rate of lexical innovation in the 18th century seems low by comparison with the earlier period, when the move towards printing in English created a gap in learned vocabulary that needed to be filled, and the 19th century, when the sheer pace of scientific progress demanded new words for new discoveries. Even in the “conservative” 18th century, it was acknowledged that everything in nature must have a name, so innovations which filled gaps in the vocabulary and which added to the clarity of nomenclature were accepted. Many of these were introduced via works such as Chambers’ Cyclopedia (Chambers et al. 1753): a search of OED online for all words with first citations from 1753 yields a total of 538, 273 or just over 50% of which were first cited in that year’s supplement to the Cyclopedia. Not surprisingly, these were mainly learned, scientific words with Latin and/or Greek etymologies. Examples are: archivist, cotoneaster, eczema, hydrangea, linguiform, phosphorical and trifoliate. Along with these classical forms, Chambers introduced words from “exotic” languages to describe the findings of explorers and plant hunters, and eponyms which name the object after its discoverer or inventor. Examples of the former are jacaranda and jacare (an alligator), both from Tupi, whilst the eponyms include camellia (named after the Moravian Jesuit botanist Kamell) and fuchsia (from the German botanist Fuchs). There are also examples of compounds, such as boat-fly, bull-fight, and butter-nut and of introductions from French, such as ballon. These trends continued in the 19th century, but on a much larger and wider scale. The vast majority of lexical innovations in this century were scientific terms formed from Latin and Greek elements, such as the medical terms conjunctivitis, myelitis, and synovitis, the botanical terms bicrenate, bifoliate, brachiate, and campylotropous, and the less technical abnormal (replacing French anormal), intensifier, paraffin, and revolver, all of which are first cited in 1835. The influx of classical formations became a cause for complaint, just as it had during the “inkhorn wars” of the 16th century. R. Chenevix Trench, one of the founders of the OED, wrote of such scientific terms: These, so long as they do not pass beyond the threshold of the science for whose use they were invented, have no proper right to be called words at all. They are a kind of shorthand, or algebraic notion of the science to which they belong, and will find no place in a dictionary constructed upon true principles (Trench 1870: 120–121).
New fields of science opened up in the course of the 19th century, requiring their own technical vocabularies. In the field of geology, newly-discovered minerals were often formed by adding the suffix -ite to the name of the discoverer or the place where they were first found. Examples are bromlite, lanarkite, leadhillite, proustite, smithsonite, stromeyerite, troostite, uralite, and voltzite, all first cited in 1835. Likewise the -itis suffix found in conjunctivitis, myelitis, and synovitis was to become very productive as a means of
5 Periods: Late Modern English naming inflammations of various body parts. Thus the scientific vocabulary snowballed in this century as the proliferation of terms provided new building-blocks for future innovation. The 19th century was also, of course, the era of British colonial expansion and imperialism. This is reflected in the wide range of sources from which words were taken in this century: just as British plant-hunters, archaeologists, and explorers plundered the world’s resources and brought them home, so the names of these trophies brought elements from “exotic” languages into the vocabulary. Examples, again all first cited in 1835, from every continent outside Europe are kiwi, rata, and tui from Maori; chacma from the African language Nama; fulwa, the Bengali word for the butter tree; nandu from Tupi/Guarani; and tepee from Lakhota. I have only been able to give a flavor of lexical innovation in the Late Modern period here: much more detail can be found in Dossena, Chapter 55. The most important point to be borne in mind is that our knowledge of lexical innovation in this, and indeed any, period, can only be as extensive and accurate as the sources available for our research. As the third edition of the OED reaches completion, we can look forward to new research findings in this area.
7 Normative grammarians In this section, I shall give a brief overview of research on what, to earlier scholars, was the main area of interest in Late Modern English: the rise of prescriptivism in the 18th century and the effect of normative texts on standardizing and regulating the language. As I have indicated above (Section 2), the seminal text in this field was S. A. Leonard’s The Doctrine of Correctness in English Usage 1700–1800 (1929). This was so influential that it led a whole generation of scholars to describe the 18th century as the age of correctness and to present 18th-century grammarians as unenlightened prescriptivists who imposed arbitrary rules on the language on the basis of logic, analogy, or conformity to the rules of Latin. In the past decade, a more nuanced view of these grammarians has come to the fore, facilitated by the wider availability of texts from this period provided by databases such as Eighteenthcentury Collections Online (Gage Cengage Learning 2009) and corpora of Late Modern texts. This has enabled scholars to scrutinize more carefully what the grammarians actually wrote and to compare their precepts with actual usage of the period. For instance, the myth that Robert Lowth “invented” the rule that a preposition should not appear at the end of a sentence still endures in cyberspace, but Ya´n˜ez-Bouza has demonstrated that he did no such thing. After examining a wide range of statements from grammarians on this topic she concludes that “Lowth was neither the first one, nor (ipso facto) was he the only one, nor was his stricture proscriptive” (Ya´n˜ez-Bouza 2008: 277). The role of prescriptive grammarians in relegating constructions such as the “double negative” to non-standard usage in this period has also received attention. Greenwood’s statement that “two Negatives or two Adverbs of Denying, do in English affirm” (Greenwood 1711: 160) is often cited as an example of mathematical logic inappropriately applied to language. Yet Nevalainen and Raumolin-Brunberg (2003) demonstrate that multiple negation was already subject to social stratification in the Early Modern period, its use largely confined to the lower classes. Thus, Greenwood may well have been rationalizing a stigma that already existed, rather than imposing an arbitrary rule on the language. Research in this area is ongoing, but the collection of papers in
75
76
I Periods Tieken-Boon van Ostade (ed.) (2008) provides a good introduction, and a salutary riposte to Leonard (1929). (See further Percy, Chapter 63.)
8 Summary This brief overview of scholarship in the field of Late Modern English has demonstrated that this is a young and growing area of research. The ever-increasing availability of corpora and electronic editions of texts from this period has led to an upsurge of interest from scholars in a number of fields. Their work challenges the previously received view that the Late Modern period was a time of relative linguistic stasis and normative attitudes to language and reveals the “complex analogical relationships” predicted by Strang (1970: 79).
9 References 9.1 Online resources Denison, David, Linda van Bergen and Joana Soliva. 2002. The English Language of the North-west in the late Modern English Period: A Corpus of late 18c Prose. http://www.llc.manchester.ac.uk/temp/ lel/david-denison/corpus-late-18th-century-prose/ Gage Cengage Learning. 2009. Eighteenth Century Collections Online (ECCO). Gale Digital Collections. http://mlr.com/DigitalCollections/products/ecco/ Nevalainen, Terttu, Samuli Kaislaniemi, Mikko Laitinen, Minna Nevala, Arja Nurmi, Minna Palander-Collin, Helena Raumolin-Brunberg, Anni Sairio (ne´e Vuorinen), and Tanja Sa¨ily. 1998–2006. Corpus of Early English Correspondence Extension (CEECE). Department of English, University of Helsinki. http://www.helsinki.fi/varieng/CoRD/corpora/CEEC/ceece.html Simpson, John (ed.). 2000–. The Oxford English Dictionary. 3rd edn. online. Oxford University Press. www.oed.com/
9.2 Printed resources Bailey, Richard W. 1996. Nineteenth-century English. Ann Arbor: University of Michigan Press. Barber, Charles. 1976. Early Modern English. London: Deutsch. Batchelor, Thomas. 1809. An Ortho¨epical Analysis of the English Language. London: Didier and Tebbett. Beal, Joan C. 1999. English Pronunciation in the Eighteenth Century: Thomas Spence’s “Grand Repository of the English Language” (1775). Oxford: Clarendon Press. Beal, Joan C. 2000. From Geordie Ridley to Viz: Popular literature in Tyneside English. Language and Literature 9/4: 343–359. Beal, Joan C. 2004. English in Modern Times 1700–1945. London: Arnold. Beal, Joan C. 2008. Shamed by your English? The market value of a “good” pronunciation. In: Joan C. Beal, Carmela Nocera, and Massimo Sturiale (eds.), Perspectives on Prescriptivism, 21–40. Bern: Peter Lang. Beal, Joan C. 2009. Three hundred years of prescriptivism (and counting). In: Ingrid Tieken-Boon van Ostade and Wim van der Wurff (eds.), 35–66. Beal, Joan C. 2011. Levelling and enregisterment in Northern dialects of Late Modern English. In: David Denison, Ricardo Bermu´dez-Otero, Chris McCully, and Emma Moore (eds.), Analysing Older English 126–139. Cambridge: Cambridge University Press. Bloomfield, M. W. and L. Newmark. 1963. The English Language: A Historical Introduction. Westport, CT: Greenwood.
5 Periods: Late Modern English ´ lvarez, Javier Pe´rez-Guerra, and Esperanza RamaBueno Alonso, Jorge L., Dolores Gonza´lez A Martı´nez (eds.). 2007. “Of Varying Language and Opposing Creed”: New Insights into Late Modern English. Bern: Peter Lang. Chambers, Ephraim, John L. Scott, and John Hill (eds.). 1753. A Supplement to Mr. Chambers’s Cyclopedia: or Universal Dictionary of Arts and Sciences. London: printed for W. Innys and J. Richardson et al. Curme, George O. and Hans Kurath. 1931. A Grammar of the English Language. Boston: D. C. Heath & Co. De Smet, Hendrik. 2007. For … to infinitives as verbal complements in Late Modern and Presentday English: Between motivation and change. English Studies 88(1): 67–94. Denison, David. 1998. Syntax. In: Romaine (ed.), 92–329. Dossena, Marina and Charles Jones (eds.). 2003. Insights into Late Modern English. Bern: Peter Lang. Finkenstaedt, Thomas, Ernst Leisi, and Dieter Wolff. 1970. A Chronological English Dictionary: Listing 80,000 Words in Order of their Earliest Known Occurrence. Heidelberg: Carl Winter. Freeborn, Derek. 1992. From Old English to Standard English. 1st ed. London: Macmillan. Go¨rlach, Manfred. 1999. English in Nineteenth-century England. Cambridge: Cambridge University Press. Greenwood, James. 1711. An Essay Towards a Practical English Grammar. London. Jones, Charles. 1989. A History of English Phonology. London: Longman. Jones, Charles. 2006. English Pronunciation in the Eighteenth and Nineteenth Centuries. Basingstoke: Palgrave Macmillan. Lass, Roger. 1999. Introduction. The Cambridge History of the English Language. Vol. III. 1476– 1776, 1–12. Cambridge: Cambridge University Press. Leonard, Stirling A. 1929. The Doctrine of Correctness in English Usage 1700-1800. Madison: University of Wisconsin Press. Lowth, Robert. 1762. A Short Introduction to English Grammar. London: R. & J. Dodsley. MacMahon, Michael. 1998. Phonology. In: Romaine (ed.), 373–535. Mair, Christian. 2002. Three changing patterns of verb complementation in Late Modern English: A real-time study based on matching text corpora. English Language and Linguistics 6(1): 105–131. Morris, Marmeduke C. F. 1892. Yorkshire Folk-Talk: with Characteristics of those who Speak it in the North and East Ridings. London: Henry Frowde. Mugglestone, Lynda. 1995. “Talking Proper”: The Rise of Accent as a Social Symbol. Oxford: Clarendon Press. Nevalainen, Terttu and Helena Raumolin-Brunberg. 2003. Historical Sociolinguistics: Language Change in Tudor and Stuart England. London: Longman. Newton, Isaac. 1687. Philiosophiae Naturalis Principae Mathematica. London: Jussu Societatis Regiae ac typis Josephe Streatii. Onions, C. T. (ed.). 1964. The Shorter Oxford English Dictionary on Historical Principles. 3rd edn. Revised with addenda. Oxford: Clarendon Press. Porter, Roy. 2000. Enlightenment: Britain and the Creation of the Modern World. London: Penguin. Poutsma, Hendrik. 1928. A Grammar of Late Modern English. For the Use of Continental Students. 2nd edn. Groningen: Noordhof. Rohdenburg, Gu¨nter. 2007. Functional constraints in syntactic change: The rise and fall of prepositional constructions in early and Late Modern English. English Studies 88(2): 217–233. Romaine, Suzanne. 1982. Socio-historical Linguistics: Its Status and Methodology. Cambridge: Cambridge University Press. Romaine, Suzanne (ed.). 1998. The Cambridge History of the English Language. Vol. IV. 1776– 1997. Cambridge: Cambridge University Press. Rose, Michael E. 1985. Society: The emergence of urban Britain. In: Christopher Haigh (ed.), The Cambridge Historical Encyclopedia of Great Britain and Ireland, 276–281. Cambridge: Cambridge University Press.
77
78
I Periods Sheridan, Thomas. 1780. A General Dictionary of the English Language. London: R. & J. Dodsley, C. Dilly and J. Wilkie. Smith, William. 1795. An Attempt to Make the Pronunciation of the English Language More Easy to Foreigners. London: T. Gillet. Strang, Barbara M. H. 1970. A History of English. London: Methuen. Tieken-Boon van Ostade, Ingrid. 2009. An Introduction to Late Modern English. Edinburgh: Edinburgh University Press. Tieken-Boon van Ostade, Ingrid (ed.). 2008. Grammars, Grammarians and Grammar-writing in Eighteenth-century England. Berlin/New York: Mouton de Gruyter. Tieken-Boon van Ostade, Ingrid and Wim van der Wurff (eds.). 2009. Current Issues in Late Modern English. Bern: Peter Lang. Trench, R. Chenevix. 1870. History in English Words. 7th edn. London: Macmillan. Walker, John. 1791. A Critical Pronouncing Dictionary. London: G. G. J. and J. Robinson and T. Cadell. Webster, Noah. 1789. Dissertations on the English Language. Boston: Isaiah Thomas and Company. Webster, Noah. 1828. An American Dictionary of the English Language. New York: S. Converse. Wright, Joseph. 1898–1905. The English Dialect Dictionary. London: Henry Frowde. Ya´n˜ez-Bouza, Nuria. 2008. Preposition stranding in the eighteenth century: Something to talk about. In: Tieken-Boon van Ostade (ed.), 251–277.
Joan C. Beal, Sheffield (UK)
6 Periods: Contemporary English 1 2 3 4 5 6 7
Introduction Phonology Morphology Vocabulary Syntax Summary References
Abstract This chapter sketches current changes in phonetics-phonology, morphology, vocabulary, and syntax, focusing on data from British, New Zealand, and Australian English. Central themes are the interplay of spoken and written, standard and non-standard English, changes in linguistic practices as opposed to changes in phonology, grammar, and vocabulary, the many changes affecting all geographical varieties of English versus changes confined to particular varieties. The changes in vocabulary illustrate the usual close connection between language and cultural change. Many of the apparent changes in grammar consist of old constructions hitherto used only in speech but now spreading into written texts. They raise questions of analysis bearing on classification (middles or passivals?), constituent structure, and particular areas of grammar, such as the tense-aspect system and modality. As usual, lying behind the changes in language and practices are issues of social and cultural capital. Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 78–96
6 Periods: Contemporary English
1 Introduction 1.1 Content This chapter presents an overview of changes in progress in English. Section 1 discusses general issues, Sections 2, 3, 4, and 5 deal with, respectively, changes in phonology, morphology, vocabulary, and syntax. Section 5 on syntax, morpho-syntax and lexico-syntax is much longer than the others. One reason is that anyone looking for further information about the phonology or morphology of any variety of English will find it in Kortmann and Schneider et al. (2004–08). Another reason is that the author of this chapter has for many years worked on the syntax of English, spoken and written, standard and non-standard and can comment on constructions not usually mentioned in corpus-based work. He also believes that syntax raises most questions about changing norms and language practices, the relationship between spoken and written language, what counts as standard and non-standard, and whether perceived changes involve new structures or a wider domain for already existing ones.
1.2 Changes in grammar and changes in language practices The question of ongoing change in contemporary English is complex, encompassing changes in the grammatical code and in language practices. One example of change in language practices is what Mair and Leech (2006: 336) call “colloquialization”, the use in written language of features associated with spoken language: semi-modals (e.g. you want to do it this way), the get passive, that or zero relative clauses, and singular they. None of these is new; in fact the last three are attested in the early 17th century. Some of them have their own functions and meanings: singular they enables writers to avoid the clumsy he or she, the get passive is dynamic and is frequently, but by no means exclusively, used for adverse events, and that is the norm in restrictive relative clauses in American English. Some of the changes presented in Section 5 relate to the use in writing of structures common in spoken English and might be counted by Mair and Leech as colloquialization (N.B. “written English” and “spoken English” are convenient simplifications; there are many genres of both). However, these structures are found in both formal and informal spoken English, and raise various questions: How old are they? What is their constituent structure? In what types of written and spoken text are they found? Which categories of users use them? These are important questions but answers to them require intensive research and detailed discussion and cannot be accommodated here. These spoken English structures are central to Section 5 because they are used by speakers of both standard and non-standard English and in combination with both standard and non-standard syntax. A major question is the extent to which regional, nonstandard norms will be followed by speakers and writers producing more formal, public texts other than dialogue in novels and plays. Much informal Internet writing already follows the norms of informal speech, norms that may be standard or non-standard but which all relate to spoken, especially spontaneous spoken, English and represent colloquialization.
79
80
I Periods
1.3 Recognizing recent or on-going changes Linguistic practices and usages are changing but typically involve the extension of existing constructions and words. Mair and Leech (2006) warn that long-term variability can be mistaken for change, that apparent changes may simply be the abandoning of “marginal shibboleths” (less books vs. fewer books), and that long-term changes may be missed. Denison (1998: 95) points out that an aberrant usage might become a successful change, might be adopted but not generally, or might remain an error not established in anybody’s usage. Denison also warns us that apparently new usages might be old ones previously confined to informal spoken English, standard or non-standard, but now appearing in formal speech and in writing. Denison’s point is supported by the examples from Mair and Leech (2006) listed in the opening paragraph, which have long been typical of spontaneous speech. A further point, emphasized in Miller (2006), is that it is difficult to classify constructions as general spoken or regional non-standard. Some apparent changes might simply be once-general usages that were restricted for a time and are now being reintroduced. (See the discussion in Section 5.1.1.) The use in British English (BrE) sat/stood instead of sitting/standing illustrates the difficulty. Carter and McCarthy (1997: 34) attribute the utterance the pilot was sat in one of the seats to “Yorkshire dialect”. In contrast, Cheshire et al. (1993: 70–71) show that BrE sat/stood is widespread and characteristic of “a general non-standard or semi-standard variety of English”. Trask (2001) reflects the uncertainty, describing was sat as “colloquial British English” and was stood as “regional British English”. Burchfield (1981), writing for the BBC, declared was sat/ stood there unacceptable in any circumstances, but almost thirty years on, the structure is used by, e.g., reporters on the BBC News at Ten (though not by the presenters). Not only is it possible that many structures considered “non-standard” are actually spoken standard, but changes in formal spoken norms and practices may be evolving much more quickly than previously thought.
1.4 The data The syntactic and morpho-syntactic data in this chapter come mainly from the author’s own database, collected over the past thirty years. Some examples were collected on the hoof from informal conversation, university meetings, and radio and television programs (some of which were recorded); some are from transcriptions of conversation and business meetings; and there are written examples from newspapers, e-mails, minutes of meetings, students’ dissertations and examination answers, and books. The discussion focuses on British and New Zealand English (NZE), which the author has experienced directly, but with supplementary data from the Australian component of International Corpus of English (ICE; for information, see http://ice-corpora.net/ice/) in the Macquarie Archive and the 4-volume Varieties of English (Kortmann and Schneider et al. 2004–8). With respect to syntax, the BrE and NZE data are typical of many other varieties of English. Many recent discussions are based on exhaustive computer searches of corpus data, which are invaluable for information about lexical collocations and morpho-syntax but not necessarily about syntax, especially the syntax of spoken language (see the discussion in Miller and Weinert 1998: Chapter 1). The trained specialist with native or
6 Periods: Contemporary English near-native knowledge of English and handling attentively all sorts of spoken and written texts on a daily basis can pick up many clues as to variant usages and possible ongoing changes. Hypotheses about changes can then be checked against corpus data (which may be behind the times) or via elicitation tests. The data presented here can confidently be taken as representing on-going changes in linguistic code or linguistic practice because of the time span, the range of speakers and writers and text types, and comparisons with other databases.
2 Phonology All geographical varieties of English evince rich phonological variation and change. Some changes are confined to particular varieties while others affect many. One example of the former is the Northern Cities Shift, six changes affecting the vowel system in the variety spoken in Chicago, Buffalo, Cleveland, and Detroit (Gordon 2008; Labov 1994, 2001). A second example is the NZE vowel in dress, which has become as close as the vowel in fleece, whereas in Australian English (AustrE) the dress vowel has become more open. The NZE vowel in trap has raised and fronted and the kit vowel has become backed and lowered (kit is allegedly perceived by Australians as cut). In modern Received Pronunciation (RP) the dress vowel has become more open, the trap vowel less forward and more open, the price diphthong more central and the square vowel a long Cardinal 3 and not a diphthong, according to Upton (2008: 241), who also mentions the falling together of paw and poor. For NZE, Bauer and Warren (2008) mention the on-going merger of the near and square diphthongs. For AustrE Horvath (2008: 105) lists ten changes in the vowel system between the 1960s and the 1990s. Most vowel changes are variety-specific, but one change affecting most if not all varieties of English is the fronting of the foot vowel. A widespread change affecting consonants is l-vocalization, observed in London and South East England, and potentially causing pairs such as meal and mill and pool and pull to fall together (Altendorf and Watt: 2008). The same l-vocalization after front vowels and in syllabic /l/ has been studied in Glasgow by Stuart-Smith et al. (2006), who note that the new vocalization exists alongside the enduring results of the old Scots l-vocalization after back vowels which gave ba’ for ball. L-vocalization is widespread in AustrE (Horvath: 2008) and is present in NZE, where it is associated with the neutralization of the vowels in pairs such as fill and full and full and fool. Another widespread change is t-glottalisation, occurring throughout the UK, as attested by the articles in Foulkes and Docherty (1999), even among speakers of modern RP (see Fabricius 2002). For varieties in London and the South East, Altendorf and Watt (2008: 212) refer to a “dramatic rise” in the frequency of /r/ realized as a labio-dental approximant in the speech of young, working-class speakers. A prosodic change affecting all varieties of English is the appearance of the High Rising Terminal (HRT). Horvath (2008) and Bauer and Warren (2008) comment on AustrE and NZE, where it is a feature of many speakers under forty. By 1990 the HRT was spreading among speakers of British English under thirty, and it is thoroughly established in the USA and Canada. The vowel mergers mentioned above are phonological in nature, but changes such as the fronting of the foot vowel, t-glottalization, and the realization of /r/ as a labio-dental
81
82
I Periods approximant are phonetic. Also phonetic, but altering the phonotactics of the affected varieties, are changes in NZE (and observed elsewhere, Laurie Bauer, p.c.) such as combinations of palato-alveolar affricate plus /r/ in try and dry, and the use of an initial palato-alveolar fricative in words such as strong and stripe. It is tempting to see some of the NZE and AustrE changes as part of a general differentiation from BrE, as is happening with vocabulary (see Section 4).
3 Morphology Observers can find many changes in linguistic practices and new uses for old forms, but recently created forms are rare apart from new plurals and singulars. Collins and Peters (2008: 342) report that standard AustrE has the past tense forms shrunk and sunk, which they consider unacceptable. However, sunk is an old form that still occurs in speech and writing in BrE and both forms are widely used in North America (Laurel Brinton, p.c.). Users of AustrE are not inventing new past tense forms but favoring one existing form over another. Changes in practice are affecting noun plurals, witness the fashion for plural forms such as fora, referenda, stadia, and corpora; Trask (2001: 130) says “fora is now increasingly used, especially by academics”. Another change affects adjectives used as nouns: plural forms such as renewables and sharps (in medical usage, sharp instruments, including needles) are common (but note examples of long standing like shallows and deeps and the names of mountains ranges such as The Remarkables and The Rockies; instructions on the use of washing machines refer to delicates and woolens, and cottons and linens). Some plural nouns are beginning to be used as singulars. Examples are bacteria, now regularly so used, and premises – notices in several shop windows in Edinburgh declare This premises is under CCTV surveillance. There are plausible reasons for the new uses. The singular bacterium is rarely seen outside technical literature, and a collection of internal spaces is most frequently referred to by singular nouns such as shop, office, house. A number of originally mass nouns are now also used as count nouns. Rugby players are said (not) to get a lot of ball, university administrators are interested in how much grant they receive, and a notice in the Western General Hospital in Edinburgh advises medical staff to avoid taking unnecessary bloods (equivalent to samples of blood not types of blood, unlike wines, butters, and so on). (Laurie Bauer, p.c., observes that the increase in the expressions such as amount of people vs. number of people indicates a more general change involving the concepts of count and mass.) Changes in derivational morphology typically consist of increased usage of a particular affix and its attachment to a larger set of words. Bauer (1994: 40–47) discusses the increasing use of -ee with non-passive meaning – escapee, absentee vs. payee, nominee, and its new use with nouns denoting non-human entities – advancee, controllee. Denison (2008) refers to a fashion among students at Manchester University for using the suffix -age, often humorously. He cites There was some general sleepage/chattage/faffage. Jespersen (1961: 436) cites chattage from a novel by Marjory Allingham (but since her novels are out of fashion and print, we can assume that the students’ use is an independent invention). In Britain outage began to replace power cut in the 1980s in connection with power cuts in North America but the Oxford English Dictionary (OED) supplies an example from 1900, also in relation to power cuts in the United States.
6 Periods: Contemporary English Jespersen (1961: 173–183) exemplifies different types of what he calls “reduplicative compounds”, such as those many many bodies (Shakespeare) or an old, old man (Dickens and Shelley). The construction is apparently becoming more frequent, sometimes with an intensifying meaning, as in It’s a big big fish (television presenter) or That’s a long long street. Note too Do you really really like me? (from the film Restless Natives). An interviewee on Channel 4 News in the UK, discussing poverty, said they’re not poor poor; that is, the persons referred to were not poor in an absolute sense but relatively poor, receiving less than 60% of the median wage. Nouns too combine, signalling ‘this X is (not) really something that is properly classed as an X’. Examples heard recently by the writer are It’s not a problem problem (i.e. no need to worry, it’s easily solved) and In spite of his age he’s a student student (i.e. he has signed up for courses and is genuinely doing the work).
4 Vocabulary The preface to the 2nd edition of the OED (Simpson and Weiner 1989: xxiv) distinguishes words of unquestioned “Anglicity” and words of doubtful “Anglicity”, including “local dialect and slang”, the technical terms of trades, processes, and science. As will be seen below, technical terms and local dialect are tenacious. Copious borrowing from other languages is usually considered a hallmark of English, but Minkova and Stockwell (2006: 483) suggest that “English has turned inwards to its own resources for new words and new readings”, and appeal to the fact that of the new words in English between 1963 and 1981 only 7.5% were borrowed. This general picture is supported by the Cambridge Advanced Learner’s Dictionary (Good [ed.] 2008), which analyzes newly included words. Out of 120 items, 20 are borrowings, much more than 7.5%, but, significantly, all under the rubric “food and drink”: e.g. basmati rice, blini, chorizo, nigiri. The other new words are either compounds from existing words or roots (grey water, energy efficient, biofuels, biomass) or old words with a new meaning (troll, hybrid). Many additions come from technology, especially computing and the Internet. Technical terms in business and commerce also find their way into general use. The dictionary lists brand aware, cool hunter, flick factor, and flexible working (but just missed credit crunch, toxic debts, and subprime). Many new items are compounds (see Mair and Leech’s comment [2006: 333] about a possible resurgence of the Germanic noun + noun sequences). Borrowing has not been abandoned by all users of all varieties of English. Consider NZE and the practices of certain users in Scotland. New Zealand society is establishing its own norms (see Schneider 2003). Cultural independence is signalled very clearly by vocabulary. New Zealanders use Antipodean vocabulary of English origin but, strikingly, Maori vocabulary too. Macalister (2006) demonstrates an early wave of borrowing into NZE from Maori (1841–1880) and, from 1970, a second period of borrowing. The latter coincides with a reviving Maori society and a growing interest in Maori culture among white New Zealanders, who had to break free of cultural (and economic) dependence on Britain when the latter joined the Common Market in 1973. Non-standard varieties too undergo lexical change, and Scots shows how complex the process can be. Macafee (1994) charts the serious loss of traditional Scots vocabulary among working-class Scots, whereas Hardie (1996) refers to the knowledge among professionals of Scots vocabulary in literature, the law, and local authority documents.
83
84
I Periods Against this background the Scottish Parliament, reconstituted in 1999 after a break of 292 years, has on its website documents in classic Scots, the former language of government. An excerpt from a document on language awareness in schools is given in (1). (1)
Amang the ettles o the study is: Tae speir oot the kind o wittins anent the languages o Scotland wantit … by education (www.scottish.parliament.uk/msp/crossParty Groups/groups/scots/SoP%20version%202.PDF; accessed 4 May 2008)
Whit ‘what’, maist ‘most’, mak ‘make’ are still in current, regular use among many speakers in Scotland. Not in regular use is speir ‘ask’; fairly esoteric are wittins ‘knowledge’, ettles ‘aims, intentions’, and anent ‘concerning’, though the last occurs regularly in legal documents. The authors of this and other documents on the same web site have gone back to written texts and revived vocabulary that otherwise was in disuse. To sum up, in the English-speaking world current developments in vocabulary are more complex and varied than might be thought, with lapses into disuse, revivals, borrowings and extensions of meaning. And in various locations all these developments accompany attempts by groups of speakers to use “non-standard” vocabulary in text types and contexts in which it is not accepted by the owners of social capital and cultural power.
5 Syntax 5.1 Middles The middle construction, as in Nothing drives like a Ford Falcon (New Zealand advertising slogan), is reviving and spreading. The middle is generally thought to exclude the progressive, to have a generic or habitual interpretation, and to require adverbs of manner. However, (2a, b) refer to single events and are identical in form and meaning with an example such as The story told well (1815 Jane Austen, Emma Chapter 23, paragraph 4). (2)
a. […] a 1912 Silver Ghost sold for £1.5m in California (2007 The Herald Scotland, Dec. 4) b. skylarks […] soon established throughout the country (Gill et al. 1994: Bird B)
There is controversy over what counts as an instance of the middle construction. In her exhaustive investigation Hundt (2007: 141–147) distinguishes mediopassives – nonprogressive, generic, focusing on inherent properties – from passivals, possibly progressive, definitely non-generic, and not focusing on inherent properties. The distinction is difficult to apply. She proposes that generic mediopassives focus on inherent properties by means of manner adverbs, modal verbs and negatives. She treats (3) as a middle/mediopassive, but a temporary transmission failure is not an inherent property; (4) she treats as a marginal middle because it describes an inherent property but is in the progressive. (3)
The fax may send if you tried again. (Hundt 2007: 142)
(4)
The 1971s, at only £1000 a bottle, are drinking so much better at the moment. (1998 Private Eye, Feb. 20, p. 7; Hundt 2007: 143)
6 Periods: Contemporary English The proposed distinction, more semantic than syntactic, is here disregarded. We propose a single middle construction with clear semantic and discourse properties which is reclaiming grammatical properties it possessed earlier and extending its range of lexical items. Note that the middle and the passive differ distinctly in syntax and meaning. The passive has its own passive morphosyntax, while the middle requires active verb forms. The passive allows optional agent phrases, the middle does not. The passive presupposes an agent, the middle does not. The middle presents an entity as controlling (Kemmer 1993) a given situation, but the passive does not. Bolinger (1968: 130) talks of “self-propelled” activities, as in The coffee is making, which applies neatly to a British Army news briefing in 2003: there were three bombs that didn’t guide for one reason or another some of them went short. The blame fell on the bombs (just as the child saying The vase just broke is blaming the vase, and might add all by itself ). Since the properties described above apply both to “mediopassives” and “passivals”, the distinction is not worth drawing. Hundt’s impressive range of middles is supported by the author’s modest data, 60 examples with 54 different lexical verbs from BrE and NZE showing that the middle is regularly used in speech and writing wherever a non-agent participant is perceived as the controller, as in (5). (5)
a. if the features are privative and require no value, then they simply check in the way that we have already seen […] (Adger 2003: 169) b. The lawsuit […] claims that the nano scratches “excessively during normal usage”. (2006 New Zealand Herald) c. that’s processing for us now (Scottish Gas employee referring to an invoice being prepared by the computer)
5.2 Indirect questions In Fiona asked if we were going to France the clause if we were going to France is an indirect question. It has the word order of a declarative clause, is introduced by if, and conveys the content of a question. The speaker’s direct question, Are you going to France?, is not reproduced. This is the construction typical of formal written English. An alternative construction is Fiona asked were we going to France, in which the complement clause has the subject-auxiliary inversion of direct questions. Quirk et al. (1985: 1052) say the construction is “common in Irish English and dialectally”. Denison (1998: 246), following Henry (1995), asserts that the construction is normal in Ulster English, Welsh English, recent American English, and the New Englishes. In fact, the “alternative” construction is typical of spontaneous spoken English in all varieties, but now regularly occurs in newspapers and in written texts not subject to copy-editing. It is not new; witness this example from Bleak House: I had thought beforehand that I knew its purport, and I did. It asked me would I be the mistress of Bleak House. (1852–53 Dickens, Bleak House, Part 14, chapter 44). The author’s private database contains written and spoken examples from final examination scripts, dissertations, newspapers (British and New Zealand), web sites, university meetings, and business meetings. The construction functions as the complement of verbs such as ask, find out (6a, b), nouns such as issue (7), adjectives such as
85
86
I Periods sure (8), and prepositions (9). Miller and Weinert (1998: 83) report that a sample of 14 Scottish English conversations contained 3 instances of the “classic” construction as in asked if they were going to France and 22 instances of the alternative construction. (6)
a. You have to ask why is it necessary to raise this very delicate and difficult subject [..] (2005 The Herald, Scotland, Feb. 17 [written]). b. Log on at the BBC World Service Aids site to find out how much do you know about condoms. (BBC webpage)
(7)
this issue about how are we preparing students to flow on seems to me quite important (Associate Deans Meeting, University of Auckland, [unplanned]).
(8)
No one is sure how long are the passages leading off from this centre. (1988 Taylor [travel article], Scotland on Sunday, Nov. 13)
(9)
The question centres on where did this come from? (Final Honours examination script, University of Edinburgh, June 2002)
5.3
NP-clause
(left dislocation)
Quirk et al. (1985: 1416–1417) assign the example in (10) to loose, informal speech. (10) This man I was telling you about – well, he used to live next door to me. They analyze This man I was telling you about as setting the “point of departure” for the utterance as a whole and as enabling speakers to avoid the tricky processing of clauses containing complex phrases. Speakers do avoid complex subject NPs in spontaneous speech, but many left-dislocated NPs are simple. Consider (11a–c): (11) a. this film it does give a real close-up of what goes on behind the scenes. (2007 Brook, “Talking movies”, BBC World, Apr. 19). b. “I like his [Keith Floyd’s] style of cookery,” adds Fenton’s wife Patricia. “He just throws everything in. Rick Stein – he’s only copying Floyd, isn’t he?” (2007 The Independent, Extra, Oct. 11, 2–3) c. “My youngest daughter gets embarrassed when she sees me on television,” says Stewart. “My eldest, she doesn’t mind so much […]” (2007 “Preparing to rock your world”, The Herald, Scotland, Nov. 13, p. 17 [Dr Iain Stewart being interviewed by Susan Swarbrick]) The construction can simply signal a contrast, as in (11b, c), or simultaneously ease the production of complex utterances and signal a contrast, as in (12)–(13). (12) “What struck me was that people who behaved the way my ex and I did, their children were fine, but those who made more mistakes, their children suffered more.” (2007 “Divorce doesn’t have to be a disaster”, The Herald, Scotland, Dec. 3, p. 15)
6 Periods: Contemporary English (13) “You know, it’s an amazing building. The one that was never built, that would have been even more amazing. It was going to be over 550 feet in height, an unbelievable sight”. (2007 The Tablet, Dec. 22–29, pp. 12–13 [Sir Terry Leahy, interview by Chris Blackhurst]) Three important points arise. The construction is not new; (14) is from Bleak House. Mr Jarndyce is not a vocative but is in apposition to Your cousin. (14) “Your cousin, Mr Jarndyce. I owe so much to him. Would you mind describing him to me?” (1852–1853 Dickens, Bleak House Chapter 4) The construction has a contrastive discourse function: Rick Stein, in contrast with Keith Floyd, the one that was never built in contrast with the one that was, the eldest daughter in contrast with the youngest. Finally, it raises interesting questions about the typological difference between subject-prominent and topic-prominent languages and the typology of spontaneous spoken English in contrast with the typology of formal written English (see Miller and Weinert 1998: 363–366).
5.4 Direct object
NP
+ complement clause
The King James Bible translates Matthew VI.28 as “Consider the lilies of the field how they grow”. Imitating the Greek (Miller and Weinert 1998: 362), the construction has the merit of making central the lilies of the field, as the referent of the direct object in a simple clause, and stating their relevant property in a separate complement clause. The constituent structure is not clear, though the lilies of the field and How they grow both modify consider. Other translators have been unhappy with the construction: the Revised English Bible has “Consider how the lilies of the field grow”, with the lilies of the field as subject of the complement clause and removed from the center of attention. The construction has been noticeable in spontaneous spoken English for some time; (15) was uttered in 1978 by a seventeen-year old Scottish male in Edinburgh, with no pause between religion and the damage … (15) i was brought up a catholic and i hate religion the damage it does to human people […] (Miller-Brown corpus of Scottish conversations, mbc2–m87, University of Edinburgh) Here the direct object of hate is religion and the complement clause conveying the relevant property of religion is the damage it does to human people. Example (16) is from the New Zealand component of ICE, again with no pause. (16) i can never remember any of my family how old they are The direct object of remember is any of my family and the relevant property is conveyed by how old they are.
87
88
I Periods Example (17) was uttered in November 2008 by the prosecuting counsel in a court case in Britain (The transcription of the words appeared on the TV screen and were spoken by actors). (17) Did you threaten Michael X at any time that you would have him killed? The direct object of threaten is Michael X and the clause that you would have him killed conveys, not a property of Michael X, but the content of the putative threat. A written version might be Did you threaten to have Michael X killed…? (The OED lists a superficially similar construction from the Wycliffe Bible of 1380: And he threatenyde hem, that thei schulden not seie to any man of him. However the clause that thei… is a clause of purpose. The people were threatened so that they would not say anything.) This sub-section closes with an example of the construction from writing (18). (18) everyone turned to [Stoichev] with a smile […] I remembered Rossi, how he’d listened so modestly to the cheers and speeches (2005 Kostova, The Historian)
5.5 Tense and aspect Two oppositions constitute the core of the English tense-aspect system, past vs. present and simple vs. progressive. These oppositions yield, e.g. writes vs. wrote, is writing vs. was writing, writes vs. is writing, and wrote vs. was writing. Early English originally had a simple present and a simple past (Elsness 1997). Appearing in the 14th century, the progressive came into regular use in subordinate clauses by the 18th century and by the late 20th century, in British English, had become very frequent in main clauses, especially in speech. Smitterberg (2000) demonstrates that the progressive became more frequent in the genres of Letters, Drama, Fiction, and History but less frequent in the genre of Science. The databases of Mair and Leech (2006: 323) show the progressive becoming generally more frequent but also spreading into new parts of the verb system such as the passive. Collins and Peters (2008: 346) observe the same phenomenon in AustrE and NZE. Here we focus on main clauses and the spread of the progressive to all lexical or situation aspects. Imperfectives express single actions, states, and habitual events. Examples of the progressive collected by the author from written texts in Britain and New Zealand suggest that the English progressive is acquiring imperfective traits. It occurs with stative verbs, as in (19), and is used for habitual events, as in (20). The stative progressives are familiar to the author, an older speaker, but the habitual progressives are either peculiar or unacceptable. Examples (19a–b) offer two stative verbs, understand and see, in the progressive, (19c) is both stative and generic, and (19d) is generic. (19) a. I am sorry to have to worry you again with […] X’s resubmission. However Department Y is still not really understanding what it is that X needs to do. (University of Auckland, e-mail from a committee chair) b. “And there is an older generation who are seeing NCEA as lowering the standards […]” (2007 New Zealand Listener, June 9–15, p. 23)
6 Periods: Contemporary English c. She lives in a house which is dating back 200 years. (2007 [photography program] BBC, June 17). d. […] it may be that internal linguistic factors […] are governing the choice between have to and have got to […] (Tagliamonte 2004: 43) These examples are significant; (19a, b, d) were written by contributors who care about their grammar; the e-mail was formal; the New Zealand Listener is a heavyweight periodical; academic texts are scrutinized by referees. Performance errors are unlikely. Examples (20a–d) occurred in final degree examination scripts. In examinations students have little time for planning and editing and produce constructions typical of speech or informal writing but unusual in formal writing. Examples (20a, b) are generic but have stative verbs in the progressive – precede, understand, and depend. Example (20c) has a progressive in a clause denoting a repeated event: the students repeatedly forget the new numbers. (20) a. The first vowel in [complaints] is short as it is preceeding [sic] the nasal bilabial /m/. (Final degree examination script, University of Lancaster, June 2002) b. Naturally a child is depending on his parents, or other adults to provide an environment were he can learn new words. (Final Honours examination script, University of Edinburgh, June 1983) c. The code is often changed and students are forgetting the new number (Minutes of Staff-Honours Students Liaison Committee Meeting, University of Edinburgh, Feb. 1998 [written by a 4th year student]) Examples (20a, b) may reflect a choice of perspective. In examinations students discuss examples given in the question paper and write down their analysis as it proceeds. They may use the progressive to metaphorically put their readers in the middle of on-going events. If correct, this explanation does not contradict the comments on the increasing frequency of the progressive but provides one of its causes. The simple present in main clauses is already isolated as a marker of certain texttypes such as sports commentaries and stage directions and of temporal and conditional subordinate clauses: When we go to London, we avoid the Underground; As soon as she arrives, we will have lunch; If you see them, please pass on my best wishes. The simple past, however, is in regular and frequent use, since speakers can use it or the present perfect to refer to past events. The simple past focuses on past situations while the perfect focuses on their results (Michaelis 1994). Miller (2000, 2004a, 2004b) argues that there are different sets of speakers with different usages of the perfect and simple past. Elsness (1997) demonstrates that the simple past with just for recent events and ever for experiential meaning (Were you ever in Paris?) is the majority use in American English, but this is also the regular pattern in the spontaneous speech of many, perhaps most, speakers in Britain. To complicate matters, there is evidence that speakers who use the perfect are beginning to combine it with specific past-time adverbs, a usage not recognized by reference grammars of standard English but mentioned in Comrie (1985: 33). (Bauer [1989] notes the same development in NZE.
89
90
I Periods He observes [p.c.] that police in New Zealand and Australia regularly use perfects in their reports to the media, even with definite time adverbs, as in an accident has occurred last night.) Elsness (1997: 250) supplies examples from Shakespeare, Pepys, and Galsworthy. As so often, an apparently new usage is an old usage spreading.
5.6 (Apparently) recent constructions 5.6.1 Latinate prefixes and Germanic particles The past thirty years have seen a growing tendency in speech and writing for Germanic particles to be combined with both prefixed and prefix-less Latin verbs, such as extend out, project out, restore back, reintroduce back, reduce down, circulate round, amplify up. In (21) reduce combines with down. Re- is a prefix, in Latin, but for most speakers of English reduce has no prefix; the particle down signals the essential “down” component of the meaning. (A noticeboard in the Botanic Gardens in Edinburgh, an eminently academic context, talks of plants being restored back into their original habitat.) (21) Deglaze the pan with the wine and reduce down to approximately 2 tablespoons. (2002 Christensen-Yule and McRae, The New Zealand Chef, p. 29)
5.6.2 Modal verbs Brown and Miller (1982) report that in their Scottish English conversations must was very infrequent and had only epistemic meaning. Deontic obligation was expressed by have to and have got to, the latter used when the obligation was imposed by someone in authority or by circumstances. Have to was neutral in that respect. May was missing altogether, whether for permission or epistemic possibility. Shall and ought were missing. Tagliamonte (2004), analyzing data collected in York (Britain) in the late 90s, reports that only 15% of the deontic forms are of must, and that must is hardly used by speakers under 30 and mostly by speakers over 70. This supports Brown and Miller’s finding, but with respect to have (got) to. Tagliamonte says that grammaticality judgments confirm that there is little to choose between I have to go shopping and I’ve got to go shopping. Grammaticality judgments are not the strongest of tests; Brown and Miller used other elicitation techniques and it would be interesting to see the results of any similar investigation in York. One major change affecting all varieties of English in the past thirty years is the use of may instead of might for the expression of remote possibility. (As a result, may has reappeared in Scottish English – see the first paragraph of Section 5.6.2.) Example (22) is spoken NZE, cited in the New Zealand Herald, and (23) is from a book by one of the UK’s leading historians, Norman Davies. (22) A St John medical adviser acknowledged that Mr Boonen may have lived had he not waited for the ambulance. (2007 New Zealand Herald, Feb. 28) (23) a. The witness said that days later he heard media reports that someone had gone missing. He believed what he had seen that night may be connected and he and his wife met with police.
6 Periods: Contemporary English The trial, before Judge Lord Matthews, continues. (2009 The Herald, Scotland June 23, p. 3) b. If the style had matched the content, it may have had more success in crossing the sectarian divide. (1999 Davies, The Isles: A History, pp. 512–513) Trask (2001:183) declares that Standard English absolutely requires might but the change is well-established in speech and writing.
5.6.3 Argument structure The argument structure of transfer verbs is susceptible to change, that is, verbs denoting the movement of something to someone and involving a person transferring, the entity transferred (typically inanimate), and the recipient. The recipient is typically human and speakers tend to place other humans at the center of events. The changes in argument structure all involve a change whereby recipient NPs are not (possibly optional) oblique objects but obligatory direct objects central to the clause syntax. The syntactic centrality parallels the semantic centrality of recipients. Verbs such as donate, attribute, and forward take the construction V NP1 to NP2, as in donated the treasure to the museum, attributed the painting to Raphael and forwarded the letter to Susan. Current dictionaries show these verbs excluding the construction V NP2 NP1, as in donated the museum the treasure. Over the past twenty years, however, each set of students taking the author’s courses has contained two or three individuals for whom donated the museum the treasure, attributed Raphael the painting and forwarded Susan the letter are perfectly acceptable. Verbs such as issue, confer, bequeath, parachute, circulate, confer, grant, and inflict typically take the construction V NP1 P NP2, where P is to or on, as in issue/bequeath something to someone and inflict something on someone. Data from written English, such as they issued him with a deadline to resign, indicate that these verbs are being used in the construction V NP2 with NP1, as in (24)–(26). (24) PM Thaksin Shinawatra is facing a showdown with critics after they issued him with a deadline to resign […] (2006 New Zealand Herald, Mar. 24) (25) It is odd that the whiskered traffic managers [of the railway companies] have bequeathed Scotland with her most pronounced token of nationhood […] (1993 Clarke, “A sporting chance for Scottish Tories” The Scotsman, Aug. 13) (26) – if it’s a severe winter we’ll have to parachute the island with food supplies like Bosnia (1994 Scotland on Sunday, Apr. 24) A passive example such as they will be conferred with honorary degrees (The Herald, Scotland) suggests the active construction confer someone with a degree. It is possible that the passive construction appeared first and gave rise to the active. Denison (1998: 215) cites a passive example from 1917 I’m issued with her [a horse]. For many speakers the pair rob and steal are clearly distinguished; you steal valuables by taking them away illegally, but you rob the owner of the valuables (Trask 2001: 250). Rob is now being used by many speakers for the action applied to the
91
92
I Periods valuables, as in the sentence They’ll probably end up going to jail or something or probably robbing stuff (BA dissertation, University of Lancaster, 2002). And of course Trask’s entry on rob and steal indicates that the usage is frequent enough to catch the attention of conservative speakers. In fact, this use of rob is an old usage that has survived in speech and in non-standard English. In Johnson’s Dictionary the third definition for rob is “to take away unlawfully”. Examples are fashion a carriage to rob love from any (Shakespeare) and Double sacrilege […] to rob the relick and deface the shrine (Dryden). Prevent occurs in three constructions: We prevented Susan from marrying a monster; We prevented Susan marrying a monster; We prevented Susan’s marrying a monster. The last construction is now archaic. Stop, which is similar in meaning to prevent, also takes the first two constructions, as in We stopped Susan from marrying a monster and We stopped Susan marrying a monster. Both verbs have to do with situations in which somebody is (metaphorically) kept away from or moved away from an action. Burchfield (1996: 622) notes the three prevent constructions, as does Trask (2001: 228) but neither mentions that the construction is attracting other verbs, such as thwart in(27) and intimidate in (28). (27) Her parents, financially thwarted from education themselves, were “adamant that we would succeed […]” (2007 New Zealand Listener, June 9–15, p. 27) (28) Witnesses are being intimidated from coming forward. (2007 New Zealand Herald, Feb.) The usual constructions are, e.g. She was thwarted in her attempt to defraud the bank and The hooligans intimidated them into handing over their cash. However, from is semantically motivated; someone thwarted or intimidated into not doing something is metaphorically removed from that activity.
5.7 Syntax and the organization of text We conclude with changes affecting conjunctions and the arrangement of clauses into sentences. Quirk et al. (1985: 667–668) classify plus as a marginal preposition, but add that it “can even be used as a conjunction”, as in You can get what you want, plus you can save money. A New Zealand example is For centuries people used liquorice […] It has anti-inflammatory, anti-bacterial, expectorant and laxative properties plus it tastes good (2007 Weekend Herald, Canvas, Sept. 9). In the above examples plus is clause-initial. Example (29) contains a sentence-initial example, with plus separated from the rest of the sentence by a comma and being treated as a sentence adverb. (29) […] it avoids the ‘duplication problem’ […] Plus, the lack of ordered rules means that OT analyses are not burdened with various intermediate levels […] (Ph.D. thesis, University of Edinburgh, 2002) Although is a subordinating conjunction introducing adverbial clauses of concession. Also is an adverb which typically occurs to the left of the main verb as in She also writes
6 Periods: Contemporary English poetry, I am also writing a book. Both have begun to function as sentence adverbs, as in (30). (30) Also, since I did not have access to the original photocopies […], I was unable to establish whether certain transformations were simply caused by transcription mistakes. (BA dissertation, University of Lancaster, June 2003) The Australian English component of ICE yielded eight examples in unscripted speech, as in (31) and (32). (31) Mmm That’s so true That’s so true Also you spent enough money on drinks on Friday night I think (S1A-094(B):333) (32) Although, English has been the most successful language in an attempt of [sic JM] becoming a lingua franca. (Final Honours examination script, University of Edinburgh, June 2001) We conclude this section with a change and a non-change. Quotative like is relatively new, at least in British English. It is absent from the corpus of Scottish English collected by the author and Keith Brown in 1977–1978 (Brown and Miller 1982) but abundant in the corpus of Scottish English collected by Jane Stuart-Smith et al. (1996) and in the corpus of English English that appears in Carter and McCarthy (1997). An excellent account of previous work and analysis of original data is offered by Buchstaller (2004). The non-change is the use of the discourse marker like. Contrary to Anderwald (2008: 459–460), it was not imported to Britain from the United States and is not recent. Utterance-final like occurs in dialogue in works by James Hogg and Sir Walter Scott from the early 19th century and utterance-medial like occurs in a mid-19th century example in Grant and Main-Dixon’s Grammar of Scots. D’Arcy (2007: 401) reports the occurrence of utterance-initial and -medial like in recordings of New Zealand speakers made in 1946–1948. Their parents were from various parts of the UK, not from the USA. Although not a new usage, the discourse-marker like deserves to be taken seriously in view of its longevity, its widespread use and the lack of detailed knowledge about its origin (see Miller 2009).
6 Summary The phonological changes described in Section 2 pertain to issues in theories of linguistic change such as the interplay between social networks, social and national identities, and properties of language systems. Space does not permit a discussion of these issues (but see Labov 1994, 2001). The lexical changes mentioned in Section 4 exemplify wellknown phenomena: creation of new vocabulary for new phenomena, (metaphorical) extension of meaning as in troll and grey water, and so on. Just as importantly and interestingly, in some varieties of English, changes in vocabulary signal the assertion or re-assertion of cultural independence. In contrast, the syntactic changes covered in Section 5 raise familiar and unfamiliar issues. Grammaticalization is exemplified by plus, which has acquired the more grammatical function of conjunction, and by both the quotative and the discourse marker like. The relation between rob and steal exemplifes the
93
94
I Periods phenomenon of persistence, whereby old structures persist in use alongside new structures to which they have given rise. The revival of the middle raises questions about transitivity in English and about basic and non-basic constructions. Context is important, of course. Mair and Leech (2006) comment that whom is far from dead; it thrives in formal texts, especially written. Shall, which few speakers of Scottish English use in speech or writing, thrives in formal notices such as This shop shall be open on Thursday evening till 9 pm (the notice was not a legal document issued, say, by the City Council but a notice produced by the shop in question). Constructions that raise a number of questions are the indirect question (Section 5.2), the NP-clause (Section 5.3), and the direct object + complement clause (Section 5.4). The constituent structure of the last construction is not clear. Dealing with the NPclause construction, one analysis puts the NP outside the clause (and has no sentences), while another one has the NP and clause as constituents of a single sentence. The constructions are unusual, too, in not fitting the theory that changes happen from below, or filter into standard English from non-standard varieties. Speakers (and writers) of standard and non-standard English alike use the constructions. The constructions raise questions about the typology of (spontaneous) spoken English, as discussed in Miller and Weinert (1998: 363–366), and about social capital and norms – who decides them and attitudes to them in informal everyday practice, in schools, and in employment. Acknowledgments: Many thanks to Laurie Bauer for his comments on an earlier draft of Section 2 in particular but also the whole chapter.
7 References Aarts, Bas and April McMahon (eds.). 2006. The Handbook of English Linguistics. Oxford: Blackwell. Adger, David. 2003. Core Syntax: A Minimalist Approach. Oxford: Oxford University Press. Altendorf, Ulrike and Dominic Watt. 2008. The dialects in the south of England: Phonology. In: Kortmann and Upton (eds.), 194–222. Anderwald, Lieselotte. 2008. English in the Southeast of England: Morphology and syntax. In: Kortmann and Upton (eds.), 440–462. Bauer, Laurie. 1989. The verb have in New Zealand English. English World-Wide 10(1): 69–83. Bauer, Laurie. 1994. Watching English Change. London/New York: Longman. Bauer, Laurie and Paul Warren. 2008. New Zealand English: Phonology. In: Burridge and Kortmann (eds.), 39–76. Bolinger, Dwight. 1968. Aspects of Language. New York: Harcourt, Brace, and World. Brown, Keith and Jim Miller. 1982. Aspects of Scottish English syntax. English World-Wide 3(1): 1–17. Buchstaller, Isabelle. 2004. The Sociolinguistic Constraints on the Quotative System: US English and British English Compared. Unpublished Ph.D. thesis, University of Edinburgh. Burchfield, Robert. 1981. The Spoken Word: A BBC Guide. London: BBC Publications. Burchfield, Robert. 1996. The New Fowler’s Modern English Usage. 3rd edn. Oxford: Clarendon Press. Burridge, Kate and Bernd Kortmann (eds.). 2008. Varieties of English 3: The Pacific and Australasia. Berlin/New York: Mouton de Gruyter. Carter, Ronald and Michael McCarthy. 1997. Exploring Spoken English. Cambridge: Cambridge University Press.
6 Periods: Contemporary English Cheshire, Jenny, Viv Edwards, and Pamela Whittle. 1993. Non-standard English and dialect levelling. In: James Milroy and Lesley Milroy (eds.), Real English, 52–96. London: Longman. Collins, Peter and Pam Peters. 2008. Australian English: Morphology and syntax. In: Burridge and Kortmann (eds.), 341–361. Comrie, Bernard. 1985. Tense. Cambridge: Cambridge University Press. D’Arcy, Alexandra. 2007. Like and language ideology: Disentangling fact from fiction. American Speech 82(4): 386–419. Denison, David. 1998. Syntax. In: Suzanne Romaine (ed.), The Cambridge History of the English Language. Vol. IV. 1776–1997, 92–329. Cambridge: Cambridge University Press. Denison, David. 2008. Patterns and productivity. In Susan M. Fitzmaurice and Donka Minkova (eds.), Empirical and Analytical Advances in the Study of English Language Change, 207– 230. Berlin/New York: Mouton de Gruyter. Elsness, Johan. 1997. The Perfect and the Preterit in Contemporary and Earlier English. Berlin/ New York: Mouton de Gruyter. Fabricius, Anne. 2002. On-going change in modern RP: Evidence for the disappearing stigma of t-glottalling. English World-Wide 23(1): 115–136. Foulkes, Paul and Gerry Docherty (eds.). 1999. Urban Voices: Accent Studies in the British Isles. London: Arnold. Gill, Brian, Juliet Hawkins, and Les McPherson. 1994. New Zealand Songbirds, Bird 8. Auckland: Godwit Press Limited. Good, Melissa (ed.). 2008. Cambridge Advanced Learner’s Dictionary. 3rd edn. Cambridge: Cambridge University Press. Gordon, Matthew J. 2008. New York, Philadelphia, and other northern cities: Phonology. In: Edgar W. Schneider (ed.), Varieties of English 2: The Americas and the Caribbean, 67–86. Berlin/New York: Mouton de Gruyter. Hardie, Kim. 1996. Lowland Scots: Issues in nationality and identity. In: Charlotte Hoffman (ed.), Language, Culture and Communication in Contemporary Europe, 61–74. Clevedon: Multilingual Matters. Henry, Alison. 1995. Belfast English and Standard English: Dialect Variation and Parameter Setting. Oxford: Oxford University Press. Horvath, Barbara M. 2008. Australian English: Phonology. In: Burridge and Kortmann (eds.), 89–110. Hundt, Marianne. 2007. English Mediopassive Constructions. A Cognitive, Corpus-based Study of their Origin, Spread, and Current Status. Amsterdam/New York: Rodopi. Jespersen, Otto. 1961. A Modern English Grammar on Historical Principles. Part VI. London/ Copenhagen: George Allen and Unwin and Ejnar Munksgaard. Kemmer, Suzanne. 1993. The Middle Voice. Amsterdam/Philadelphia: John Benjamins. Kortmann, Bernd and Edgar Schneider, in collaboration with Clive Upton, Kate Burridge, and Rajend Mesthrie. 2004–08. Varieties of English. 4 vols + 1 CD ROM. Berlin/New York: Mouton de Gruyter. Kortmann, Bernd and Clive Upton (eds.). 2008. Varieties of English 1: The British Isles. Berlin/ New York: Mouton de Gruyter. Labov, William. 1994. Principles of Linguistic Change. Vol. 1. Internal Factors. Oxford/Malden, MA: Blackwell. Labov, William. 2001. Principles of Linguistic Change. Vol. 2. Social Factors. Oxford/Malden, MA: Blackwell. Macafee, Caroline I. 1994. Traditional Dialect in the Modern World: A Glasgow Case Study. Frankfurt: Peter Lang. Macalister, John. 2006. The Maori presence in the New Zealand English lexicon, 1850-2000. English World-Wide 27(1): 1–24. Mair, Christian and Geoffrey Leech. 2006. Current changes in English syntax. In: Aarts and McMahon (eds.), 318–342.
95
96
I Periods Michaelis, Laura A. 1994. The ambiguity of the English present perfect. Journal of Linguistics 30: 111–157. Miller, Jim. 2000. The perfect in spoken and written English. Transactions of the Philological Society 98(2): 323–352. Miller, Jim. 2004a. Problems for typology: Perfects and resultatives in spoken and non-standard English and Russian. In: Bernd Kortmann (ed.), Dialectology Meets Typology: Dialect Grammar from a Cross-Linguistic Perspective, 305–334. Berlin/New York: Mouton de Gruyter. Miller, Jim. 2004b. Perfect and resultative constructions in spoken and non-standard English. In Olga Fischer, Muriel Norde, and Harry Perridon (eds.), Up and Down the Cline: The Nature of Grammaticalization, 229–246. Amsterdam/Philadelphia: John Benjamins. Miller, Jim. 2006. Spoken and written English. In: Aarts and McMahon (eds.), 670–691. Miller, Jim. 2009. Like and other discourse markers. In: Pam Peters, Peter Collins, and Adam Smith (eds.), Comparative Studies in Australian and New Zealand English, 317–338. Amsterdam/Philadelphia: John Benjamins. Miller, Jim and Regina Weinert. 1998. Spontaneous Spoken Language: Syntax and Discourse. Oxford: Clarendon Press. Minkova, Donka and Robert Stockwell. 2006. English words. In: Aarts and McMahon (eds.), 461–482. Quirk, Randolph, Sidney Greenbaum, Geoffrey Leech, and Jan Svartvik. 1985. A Comprehensive Grammar of the English Language. London: Longman. Schneider, Edgar W. 2003. The dynamics of new Englishes: From identity construction to dialect birth. Language 79(2): 233–281. Simpson, J. A. and E. S. C. Weiner (eds.). 1989. Oxford English Dictionary. 2nd edn. Oxford: Oxford University Press. Smitterberg, Erik. 2000. The progressive form and genre variation during the nineteenth century. In: Ricardo Bermu´dez-Otero, David Denison, Richard Hogg, and C. B. McCully (eds.), Generative Theory and Corpus Studies. A Dialogue from 10 ICEHL, 283–297. Berlin/New York: Mouton de Gruyter. Stuart-Smith, Jane, Claire Timmins, and Fiona Tweedie. 2006. Conservation and innovation in a traditional dialect. L-vocalization in Glaswegian. English World-Wide 27(1): 71–87. Tagliamonte, Sali. 2004. Have to, gotta, must: Grammaticalization, variation and specialization in English deontic modality. In: Hans Lindquist and Christian Mair (eds.), Corpus Approaches to Grammaticalization in English, 33–55. Amsterdam/Philadelphia: John Benjamins. Trask, R. Larry. 2001. Mind the Gaffe: The Penguin Guide to Common Errors in English. London: Penguin. Upton, Clive. 2008. Received Pronunciation. In: Kortmann and Upton (eds.), 237–252.
Jim Miller, Edinburgh (UK)
II Linguistic Levels 7 Linguistic Levels: Phonology 1 Introduction to some basic terms and developments in phonological theory 2 Stability and instability in consonant inventories 3 The future of looking back in time on sound changes 4 Summary 5 References
Abstract Changes in the sound system of a language may involve many different aspects. First, phonemes may be added to an inventory, they may become obsolete, or they may change their shape. Second, allophonic rules may emerge (e.g. “voice intersonorant fricatives”), disappear, or change. Third, phonotactic restrictions may be added (e.g. no syllable-initial /kn/ sequences, so that such clusters are reduced to one sound), or change their effect (e.g. the ban on /kn/ sequences may be resolved by epenthesis). Fourth, prosodic structure may change (resulting in, for instance, stress shift) and, fifth, morphophonological alternations (e.g. ablaut and umlaut) may start to play a different role or they may vanish (e.g. when morphology is regularized). This chapter will first briefly introduce the field of phonology as envisaged by structuralists and generative linguists. Section 2 discusses some changes in the consonant inventory, the allophonic variations, and the phonotactic restrictions in the history of the English language and shows that many of these changes are interrelated in the sense that a change in one component triggers an effect in another component. Section 3 expresses some ideas on the future of looking back on changes in the sound system, and Section 4 summarizes the chapter.
1 Introduction to some basic terms and developments in phonological theory The ability of speakers to distinguish separate segments in a string of speech sounds is an important part of the knowledge that language users have about their native language. Speakers of English realize that the words pin, thin, bin, and fin differ only in the first sound and they may use this knowledge in rhyme, alliteration and in games with nonsense words (e.g. “Annie the pannie the thanny”, “Simon the bimon the fimon”). The sounds represented by p, th, b, and f have a function in English, i.e. these speech sounds – together with approximately 35 others – are the minimal units that can distinguish meaning in English. The smallest elements that can cause a change in meaning are called “distinctive sounds” or “phonemes”. Jones (1967: vi) mentions that the French word “phone`me” appears to have been invented by the Frenchman L. Havet, who used it in 1876 to mean “speech sound”. At the beginning of the 20th century the term “phoneme” acquired a more abstract meaning and referred to the minimal unit in speech that can function to distinguish meaning. Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 97–113
98
II Linguistic Levels A theory of the phoneme began to be developed at the turn of the 20th century – especially in the works of Baudouin de Courtenay (1895) and de Saussure (1916) – and became one of the important research interests of structural linguistics in the first half of the 20th century. In a statement submitted to the First International Congress of Linguistics meeting in The Hague in 1928, Roman Jakobson, Sergej Karcevskij, Nikolaj Sergeevicˇ Trubetzkoy, and other members of the Prague School emphasized that a scientific description of a language must include a characterization of its phonological system, i.e. the repertory, pertinent to that language, of the distinctive contrasts among its speech sounds (see Jakobson 1971 [1962]). Thus, an important goal of structural phonology in Europe – as well as in North America – was to establish the phoneme inventories of languages. The method used to establish the phoneme inventory of a language is to systematically compare words with different meanings that differ in one sound only. Such word pairs are called “minimal pairs”. Furthermore, the international congress held that the field of language change should not be confined to studying isolated changes; rather, changes should be considered in terms of the linguistic system which undergoes them. As Waugh (1976: 21) puts it, according to the Prague School, “we must understand the structure before the change begins, the structure after it takes place, the sense of the change undergone in respect to the undergoing system, the level at which the change takes place (e.g. distinctive feature or phoneme) and the effects of the change on the system”. One of the main concerns of Roman Jakobson became to develop a theory of phonology that would predict exactly those distinctive sounds that can be found in the world’s languages, and he hypothesized that there is a limited number of phonological (or “distinctive”) features – approximately 15 – that characterize the sounds of human languages (see, e.g., Jakobson 1939; Jakobson et al. 1951; Jakobson and Halle 1956). In the system that Jakobson and his colleagues developed, each phoneme is represented by a set of features such as [grave]/[acute] and [flat]/[nonflat] which are unrelated (i.e. not grouped into smaller sets) and binary. For Jakobson, speech sounds are characterized by features and for him it followed that sound changes should involve features or phonemes too: a sound change is a change in the distribution of a phonological feature or phoneme within a system. With Sapir (1921) before him, Jakobson posited that if in a system /p, t, k/, one segment has changed (e.g. /p/ to /b/), the outcome is asymmetric (/b, t, k/). The simplest way to restore the symmetry is an analogous change of the other members in the system (e.g. /t, k/ → /d, ɡ/). Should the resulting pattern already exist, then the system of oppositions can only survive if the older series (in our example /b, d, g/) itself undergoes a change (e.g. spirantization). In this hypothetical example, the stops specified as being voiceless gradually changed into voiced stops – i.e. the feature that expresses voicing in stops changed – and the older voiced stops became fricatives – i.e. for voiced stops the feature that specified complete obstruction in the vocal tract changed into a feature that specifies incomplete obstruction ([+continuant]). In this case, the phonological sound change alters the relationship between two elements from /p/ versus /b/ to /b/ versus /v/. Jakobson also allowed for sound changes that lead to the elimination of a phonological contrast or to the formation of a contrast and we will encounter English examples of modification, elimination, and creation of phonological contrasts in the remainder of this chapter. Speech sounds may be articulated differently depending on the position in the word. For example, the initial sound in the word pin (and the non-word pannie) is pronounced
7 Linguistic Levels: Phonology with a puff of air (called “aspiration”), which is lacking in the word spin. If we were to replace the sound p in pin and pannie by the corresponding p-sound in spin or spaniel, or by the unaspirated p-sound used in Dutch pin, the words would sound odd to native speakers of English, but pin would still mean ‘a short thin piece of metal used for fastening things together’. Aspiration thus does not change the meaning of the words in question, but it is part of the grammar that speakers of Present-day English employ. The phonological rule that applies in this case would be “in word-initial position, add the feature [spread glottis] (for aspiration) to segments specified as [stiff vocal folds] underlyingly”. Speech sounds that do not cause a change in meaning, but are realizations of one sound in different contexts (e.g. aspirated [ph] in absolute word-initial position versus unaspirated [p] preceded by /s/) are called “allophones” (a term that was invented about 1934 by Benjamin Lee Whorf). To explain sound alternations in particular contexts, it is useful to distinguish the underlying representation from the actual realization (or “surface representation”): the former is put between slashes (“/ /”) and the latter is put between square brackets (“[ ]”). For example, the underlying form of the English regular past tense marker spelled -ed is /d/ (as in fail[d], love[d]). Underlying /d/ is realized with an epenthetic vowel after a stem that ends in an alveolar stop (need[ɪd], want[ɪd]) and it is realized without voicing – i.e. as [t] – after a stem-final voiceless obstruent (as in kiss[t], walk[t]). In generative frameworks, the surface form (or “output”) is derived from the underlying form (or “input”) by phonological rules and in Optimality Theory (OT), the surface form that violates the least highly ranked constraints compared to alternative surface forms is selected as the “optimal” one for phonetic realization. In generative phonology, changes in allophonic variation may thus involve a change in rule applications, whereas in OT, allophonic changes involve a reranking of universal constraints in the language specific grammar. Apart from the knowledge about which sounds are part of the sound system of a language and the knowledge about the realization of sounds in particular contexts, speakers also have clear intuitions about how sounds are organized to form words, i.e. how some sounds may combine with other sounds in the language. Speakers of English know that /p/ may combine with /s/ – as in spin – whereas the th-sound may not (*sthin). Changes in the phonotactics of the English language have occurred throughout its history and we will consider a few of them in this chapter. Some phonological phenomena have scope over larger domains than one segment, e.g. a syllable, or take effect at a morphosyntactic boundary. An example of the first kind of phenomenon is word stress (see Minkova, Chapter 8, which discusses the development of word and phrasal stress from Old to Present-day English). As an example of a phonological phenomenon that applies across a morphosyntactic boundary, English consonant-intrusion can be mentioned here. To avoid two vowels becoming adjacent, some varieties of English have the option of inserting a sonorant consonant. Which particular sound is inserted depends on the preceding vowel: the glide /j/ is inserted after a front high vowel (e.g. I see [j] it), /w/ is inserted after a back high vowel (e.g. too [w] old) and an r-sound is inserted when a non-high vowel-final word is followed by a suffix or a word that is vowel-initial (e.g. I saw [r] it), for instance in the Eastern Massachusetts dialect (McCarthy 1993: 170–171). Thus, the epenthetic consonant is a sonorant and the preceding vowel determines whether it is realized as /j/, /w/, or rhotic /r/. The present chapter is mainly concerned with segmental phonology, i.e. the study of the function, behavior, and organization of speech sounds in one language and across
99
100
II Linguistic Levels languages. The three factors mentioned immediately above are part of the knowledge that speakers have about the sound system of their native language (see, e.g., Goldsmith 1995: 1–13; Lass 1984). The problem for historical linguistics is of course that there are no speakers of earlier stages of the language alive. Fortunately, in many cases they left traces of their speech in written records (manuscripts, grammars, etc.) and where written records are absent, it is often possible to reconstruct phonological systems by comparing different languages and language varieties that derived from the language period we are interested in. The present chapter examines some changes in the phoneme inventory, the allophonic variations, and the phonotactic restrictions in the history of the English language. It is not our ambition to investigate all phonological properties; we will concentrate on only a few properties that concern speech sounds and that have changed over time in the history of the English language as spoken in England. The most well-known phonological changes in the history of English are changes in the vowel system. Since these changes are discussed at length in other chapters of this volume (e.g. the Great Vowel Shift in Krug, Chapter 48), we will refrain from repeating the story of vowel changes here and concentrate on changes in the consonant inventory, allophonic variation of consonants, and phonotactic restrictions on consonant clusters. The chapter on prosody (see Minkova, Chapter 8) complements this one and considers phenomena that cannot be restricted to single speech sounds (e.g. word and phrasal stress assignment).
2 Stability and instability in consonant inventories At the turn of the 20th century, the major goal of historical linguistics was to reconstruct phonological systems and rules of language stages of which we have no direct evidence. According to an extreme version of the Neogrammarian doctrine, historical study was the only genuinely scientific approach to the facts of language. By examining sound changes, the Neogrammarians tried to find out why modern languages have developed the way they have and how they relate to other languages. The Neogrammarians maintained that language change is systematic and takes place without exceptions. For example, it was already observed by e.g. Friedrich von Schlegel in 1801 that Latin words that start with a labial plosive (e.g. ped, pisces) correspond to words that start with a labial fricative in Germanic languages (English foot, fish; German Fuss, Fish) and that words beginning in (/t/) or (/k/) in Latin would have (/θ/) or (/x/ or /h/) in early stages of Germanic languages. The Neogrammarians deduced from these observations that the voiceless stops /p, t, k/ in the common ancestor of these languages (i.e. Indo-European) remained unchanged in Latin, but became voiceless fricatives in Germanic. By applying their method of systematically comparing sounds in certain positions of similar words in different languages, the Neogrammarians constructed family trees of related languages. Even though it is still common to refer to language family trees and say that, for instance, English, German, Dutch and Frisian belong to the “branch” of West-Germanic languages and are “sister” languages, few people today would support the idea that the origin of the languages that we now know can be traced back to one common ancestor (see e.g. Aitchison 2001: 23–36, Wunderlich 2008). Nevertheless, the important contribution of Neogrammarians to linguistic research is the insight that some sound changes do not take place arbitrarily, but that classes of sounds in a particular linguistic context may undergo a certain change.
7 Linguistic Levels: Phonology To illustrate how consonant inventories may change over time, consider as an example the inventory of obstruents that the Neogrammarians assumed for Indo-European (1a) and its descendant Proto-Germanic (1b): (1)
a. Indo-European obstruents b. Proto-Germanic obstruents /f, θ, xj, x, xw/ /p, t, kj, k, kw/ /b, d, ɡj, ɡ, ɡw/ /p, t, kj, k, kw/ /b, d, ɡj, ɡ, ɡw/ /bh, dh, ɡhj, ɡh, ɡhw/ /s, h/ /s, h/ and geminate stops after short vowels geminate stops after short vowels
What remained “stable” in the transition from the Indo-European obstruent inventory to the Proto-Germanic one is the number of phonemes: both systems distinguish 17 obstruents. What changed was the set of phonemes. Proto-Germanic has fewer laryngeal contrasts, but more contrasts in manner of articulation for the different places of articulation as a result of a set of consonant shifts – commonly referred to collectively as “Grimm’s Law” – whereby the Indo-European voiceless stops spirantized and became voiceless fricatives, the unaspirated stops underwent a strengthening process and became voiceless aspirated stops and, finally, the so-called “breathy voiced” stops were deaspirated and became voiced unaspirated stops (e.g. Harbert 2007: 41–88; for a phonological account of parts of this consonant shift see, e.g. Iverson and Salmons 1995). This type of sound change is an example of what Jakobson would consider a modification of a phonological contrast: it altered the relationship from voiceless stop to voiced stop and from voiced stop to breathy voiced stop into the opposition voiceless fricative versus voiceless stop and voiceless stop versus voiced stop. We can only speculate as to why the system changed the way it did. The internal change from voiceless stops to voiceless fricatives in one branch of a language family may perhaps be attributed to the fact that voiceless /p, t, k/ are perceptually close to voiced /b, d, ɡ/ whereas the fricatives /f, θ, x/ and the stops /b, d, ɡ/ are more distinct. One hypothesis may thus be that in order to make the two classes more perceptually distinct, one class was “enhanced” by spirantization. However, Proto-Celtic kept the /p, t, k/ versus /b, d, ɡ/ contrast and conflated (or “merged”) /b, d, ɡ/ and /bh, dh, ɡh/, so that two perceptually close classes emerged as in (2a, b): (2)
a. Indo-European obstruents b. Proto-Celtic obstruents /(p), t, (kj), k, kw/ /p, t, kj, k, kw/ j w , ɡ, ɡ / /b, d, (ɡj), ɡ, (ɡw)/ /b, d, ɡ h h hj h hw /b , d , ɡ , ɡ , ɡ / /s, h/ /s, h/ and geminate stops and geminate stops
In this case, what remained “stable” (i.e. the “pertinent” properties) are the different manners of articulation. The unstable or “transient” property is the reduction in the number of phonemes: Proto-Celtic exhibits fewer laryngeal contrasts and – due to depalatalization and delabialization – fewer places of articulation. This type of sound change is an example of what Jakobson would consider elimination of a phonological contrast. The change altered the three-way opposition from voiceless stop to voiced
101
102
II Linguistic Levels stop to breathy voiced stop into a two-way contrast between voiceless stop and voiced stop. The question why Proto-Germanic opted to enhance the contrast by spirantization of one class, whereas Proto-Celtic kept the /p, t, k/ versus /b, d, ɡ/ contrast and conflated /b, d, ɡ/ and /bh, dh, ɡh/ is notoriously difficult to answer. In the framework of Optimality Theory as introduced in the early 1990s in different works by McCarthy, Prince, and Smolensky (McCarthy and Prince 1993a, 1993b, 1995; Prince and Smolensky 1993), the suggestion is made that the desire to satisfy “perceptual distance” to accommodate the needs of the hearer was valued more in Proto-Germanic than the constraint “avoid marked elements” (i.e. fricatives are universally more marked than stops and thus disfavored in phoneme inventories). In contrast, the constraints “faithfulness to manner” and “minimize phoneme inventory” (i.e. the so-called “principle of economy”) gradually gained more weight and thus became more important than “perceptual distance” in Proto-Celtic. Of course this only says what happened in different language communities and not why it happened. As an autonomous reviewer pointed out, we cannot answer the question why “perceptual distance” was more valued in one language family and less in another and why the “principle of economy” came to play a more important role in some language communities, because we do not know the reasons for variation preferences in speech communities.
2.1 Stability and instability in consonant inventories related to phonotactics Proto-Germanic is the ancestor of Old English (OE). In (3) we compare the inventory of obstruents assumed by the Neogrammarians for Proto-Germanic and the inventory of Old English as proposed by Murray (Chapter 17; see also Marsh, Chapter 1). (3)
a. Proto-Germanic obstruents /p, t, kj, k, kw/ /b, d, ɡj, ɡ, ɡw/ /f, θ, xj, x, xw/ /s, h/ and geminate stops after short vowels
b. Old English obstruents /p, t, kj, k/ /b, d, ɡj, ɡ/ /f, θ, xj, x/ /s, ʃ, h/ and geminate consonants (except /h/, /ʃ/ and /r/) after short vowels
The effect of OE consonant gemination in intervocalic position was to create a closed syllable. Geminate consonants were lost before the 13th century. Note first of all that Murray (Chapter 17) does not assume labiovelar stops and fricatives for Old English. The spelling of OE cwe¯n ‘queen, wife of a king’ and cwæþ ‘said, spoke, called, named, proclaimed’ suggests that /kw/ was a legitimate sound or sound sequence. Rather than being analyzed as one phoneme, the spelling suggests that we are dealing with a consonant cluster of a velar stop followed by a labiovelar approximant /w/ (due to the influence of Anglo-Norman scribes, the spelling changed into in the Middle English [ME] period). Murray does assume palatalized velar stops and fricatives in the Old English consonant inventory, but these were most probably fronted realizations of the velar obstruents before front vowels.
7 Linguistic Levels: Phonology We will now focus on the fact that before the change from Proto-Germanic to Old English, there are some asymmetries in the consonant system. The ones we will have a closer look at are: (i) only velar obstruents have a labialized counterpart (e.g. /k/ contrasts with /kw/, but /t/ does not contrast with /tw/) and (ii) only alveolar stops and fricatives are followed by /w/ (cf. OE twa ‘two’, twelf ‘twelve’, twentig ‘twenty’, betwix ‘between’, and sweord ‘sword’; /w/ disappeared in some initial /tw/ and /sw/ clusters preceding a round back vowel in late Middle English and Early Modern English, so that is a “silent” letter in the words two and sword today). If it is true that the labiovelar segments were reanalyzed as consonant clusters of a velar stop followed by a labiovelar approximant at an early stage of Old English, the phoneme inventory changed in that labiovelar obstruents gradually became obsolete and the phonotactics changed as a consequence: the language now allowed alveolar and velar obstruents followed by a labiovelar approximant in the onset of a syllable. Clusters with labial obstruents followed by /w/ did not emerge, presumably due to a ban on identical places of articulation in two adjacent onset consonants, i.e. a so-called “OCPeffect” (Obligatory Contour Principle), which prohibits /pw/, /fw/, /tl/, /dl/, etc. as possible word-initial clusters. The change described above from a complex phoneme to a consonant cluster is one example of how the loss of a phoneme was compensated for by relaxing the phonotactics. The reverse state of affairs is also attested: at least in one case, a phonotactic restriction (“disallow sequences of /s/ immediately followed by /k/”) resulted in the emergence of a new phoneme. Some sequences of alveolar /s/ followed by the velar stop /k/ were at some point no longer pronounced as /sk/ by the Anglo-Saxons, but rather as the alveopalatal sound /ʃ/ (as in the words ship, sheep, shoe and fish). It is often argued that in Old English, the change from /sk/ to /ʃ/ was the result of palatalization (see below). However, palatalization occurred in the context of front vowels, whereas /sk/ clusters seem to have become alveopalatal fricatives in more environments, for instance, word-initially independent of the quality of the following vowel, word-medially (except where the cluster is not tautosyllabic before back vowels, e.g., *aisko¯jan → a¯scian ‘to ask’ where represents /sk/) and word-finally. A similar assimilation process must have applied in the German language as well, since words of the same origin – so-called “cognates” – are also pronounced with an alveopalatal fricative in this language (as in the Ger. words Schiff ‘ship’, Schaff ‘sheep’, Schuh ‘shoe’, and Fisch ‘fish’). Other West-Germanic languages such as Frisian and Dutch still have initial consonant clusters in these words (e.g. Frisian [skip] and Dutch [sxɪp] for ‘ship’). In different Germanic systems, /sk/-clusters are more or less stable. For instance, in the history of Icelandic, /sk/ remained fairly stable, whereas /sk/ was less stable in the history of English and very unstable in the history of German, where such clusters are now extremely rare. Even though these three languages derived from a common origin, they put different restrictions on consonant clusters in their respective histories. We conjecture here that Icelandic retained /sk/ clusters because they were highly frequent and could be syllabified as coda-onset clusters in most environments (e.g. fis.
103
104
II Linguistic Levels kur ‘fish-NOM’, fis.kinn ‘the fish-ACC’); they are retained in environments where other languages would ban them because of analogy (e.g. in fisk ‘fish-ACC’) – here the cluster is not modified in analogy to the form fiskur ‘fish-NOM’). In English, /sk/ clusters were modified at a particular stage in the history of the language and they were reintroduced when a large number of loanwords from Old Norse, Latin, etc. entered the language which had /sk/-clusters. Finally, in German, /sk/ clusters underwent a change that also affected most loans (presumably because German never borrowed as many words which involved /sk/-clusters as English did, so that the driving force to reintroduce them – i.e. the renewed relative frequency of /sk/-sequences – did not apply here; only a very few loanwords, all of Greek origin, retained the /sk/ cluster, e.g. Ger. Skelett ‘skeleton’). Theoretically, there are many ways to resolve the ban on /sk/ clusters, e.g. metathesis (/sk/ → /ks/), prosthesis (/sk/ → /ɛsk/), epenthesis (/sk/ → /sәk/), simplification (/sk/ → /s/ or /sk/ → /k/), gemination (/sk/ → /ss/ or /kk/), change of the place or manner of articulation of one of the consonants in the cluster (/sk/ → /sx/), or merger, so that the place of articulation of one segment survives and the manner of articulation of the other segment survives (/sk/ → /t/ or /sk/ → /ʃ/). The open question is why a language opts for which solution. The first possibility, metathesis, is not favored in word-initial positions, because the result does not make a better word-onset than the original cluster. Prosthesis is often found in a context of a preceding consonant, so that the preceding consonant can fill the onset position and the prosthesized vowel can form the nucleus of a new syllable of which /s/ can be the coda consonant. Prosthesis is attested in texts from the 2nd century onwards in word-initial /s/ plus voiceless stop clusters after consonants in Romance languages. This process gradually spread to all contexts in which word-initial /s/ is followed by a stop (e.g. Fr. esprit ‘spirit’ and Spanish estado ‘state’). English has borrowed a considerable number of words with initial /ɛs/ plus stop clusters from Romance (e.g. escape). Epenthesis is an option that second language users often employ to remedy syllable structures that do not occur in the native language. Simplification of /s/ plus consonant clusters is an option that is often found in early child speech. Kiparsky (2003: 329) points out that assimilation of consonant clusters resulting in gemination seems to happen in languages that already have geminates, whereas languages without pre-existing geminates prefer to simplify clusters. A change of manner of articulation of the second consonant (/sk/ → /sx/) has taken place in Dutch. The merger of the segments /s/ and /k/ into one that shares properties with both consonants such as /t/ – which like /s/ is alveolar and which like /k/ is a stop – is unlikely, because the single consonant /t/ does not reflect the perceptually distinct stridency of the original /sk/ cluster. In Old English, the strident nature of the first element in the cluster is preserved and the alveolar fricative is retracted in the context of the following back consonant which is left unrealized (/sk/ → /ʃ/). This process introduces a new segment: the system did not accommodate alveopalatals before the change took place and the effect of the change is thus an extension of the consonant inventory. Note that we do not need to assume the introduction of a novel distinctive feature for this change if we assume that velar sounds are specified for the feature [back]; alveopalatal fricatives may be then specified for [strident] as well as [back], so that the contrast between /s/ [strident], /ʃ/ [strident, back] and /x/ [back] can be expressed by features that were already distinctive in the phonology of speakers of English. One of the most interesting aspects of linguistic change is the fact that some units resist change. In Romance languages, prosthesis affected not only /sk/, but also /sp/
7 Linguistic Levels: Phonology and /st/. In English, only /sk/ underwent a change. The fact that /sk/ is modified whereas /sp/ and /st/ do not results in an asymmetry in the system: now only labial and alveolar stops follow /s/; velar stops no longer do. However, the change from word-initial /sk/ → [ʃ] did not apply across the board and was not permanent. Consider in this respect that under the influence of Scandinavian invaders and Latin scholars, loanwords were introduced into the language which started with /sk/ clusters and these clusters were not modified, but borrowed as such into the English language (e.g. skill and sky, which are loans from Old Norse). The temporary ban on /sk/ clusters resulted in a change in the linguistic system. The change /sk/ → [ʃ] itself was short lived, but it had a long-lasting effect. In Jakobson’s terminology, this sound change is the “formation of a contrast”: the effect of the change on the system is the creation of a new contrast within the class of fricatives (such that alveolar /s/, for instance, now contrasts with alveopalatal /ʃ/).
2.2 Stability and instability in consonant inventories related to allophonic variation In many varieties of Old English, most notably West Saxon and Northumbrian, the Germanic voiceless velar stop /k/ gradually developed into the voiceless palatal affricate [tʃ] in the following three contexts: (i) if initial followed by a front vowel or /j/ (e.g. OE cild [kild] → [kjild] → [tʃı¯ld] ‘child’; OE short vowels lengthened sometime around the 10th century if they were preceded by /ld/ clusters. Long vowels later underwent the Great Vowel Shift (e.g. OE /kı¯ld/ → Early Modern English /tʃaild/) (see Krug, Chapter 48). (ii) if medial preceded and followed by a front vowel (e.g. OE cwice ‘quitch’) and (iii) if final preceded by front vowel (e.g. OE ic [ɪk] → [ɪtʃ] ‘I’). The voiced velar stop /ɡ/ underwent a similar process: if followed by a front vowel it was realized as /j/ (e.g. *georn → [j]ern ‘eager’). The voiced stop /ɡ/ was palatalized if preceded by a front vowel (OE bricg [brɪɡ] → [brɪɡj] → [brɪdʒ] ‘bridge’), or between a nasal and /j/ (e.g. *sangjan → sen[dʒ]an ‘to singe’), or when geminated before /j/ (e.g. *laggian → lecgan ‘to lay’, where represents /dʒ/). At the time when palatalization of /k/ and /ɡ/ to [kj] and [ɡj]/[j] took place, the consonant inventory was stable and no phonological contrast was introduced. Rather, the segments /k/ and /ɡ/ had allophonic variants which they did not have before. In later stages in the history of English, we find velar stops before front vowels both in words of Germanic origin and in borrowed words. With respect to words of Germanic origin consider that for instance PGrmc. *kunningaz had a velar stop before a back vowel. At the time that palatalization applied in Old English, the velar stop was still followed by a back vowel. Later, the vowel in question was affected by a process commonly referred to as “i-mutation”, by which stressed long and short back vowels were fronted in the context of a following high front segment ([i], [j], [y]): PGrmc. *kunningaz > OE kynning (mostly written as ) ‘king’. After i-mutation, the process of palatalization did not apply and for this reason, many words of Germanic origin have a velar stop followed by a front vowel. Also note that palatalization of velar stops did not take place in Scandinavian borrowings (e.g. Present-day English kid, kettle, dike,
105
106
II Linguistic Levels give, get, egg). Thus, [tʃ] and [dʒ] are not allophonic variants of the phonemes /k/ and /ɡ/ in Modern English. Instead, after the rule of palatalization ceased to play a role in the phonology of Old English and after borrowings from Scandinavian and other languages and after the change from [kj, gj] to [tʃ, dʒ], the consonant inventory is expanded, i.e. a new system emerged in which the phonemes /k/ and /ɡ/ contrast with the phonemes /tʃ/ and /dʒ/, respectively. (For a more detailed account of palatalization, the reader is referred to Murray, Chapter 17, Section 3.3.) There is a lot of debate over the realization of velar fricatives. Most textbooks maintain that OE /x/ was realized as [h] word-initially before vowels and before the sonorants /n, l, r, w/, it was pronounced as [x] medially and finally except after front vowels and it was realized as palatal [c¸] after front vowels. The variant [c¸] gradually vocalized in all southern English dialects, resulting in compensatory lengthening of a preceding short vowel, which was later affected by the Great Vowel Shift (e.g. [ɪc¸] → [iː] → [ai] in words like knight, and night). In late Middle English, the variant [x] was labialized and changed into [f] when following round vowels (mostly /u/) in some dialects (cf. ModE enough where = /f/) and gradually disappeared entirely in other phonological contexts. With respect to the direction of phonological change, it is interesting to reflect on the following developments: in some variants of English, /xt/ changed into /ft/ after a short back vowel (cf. Modern English draught), whereas in the early history of Dutch, monomorphemic /ft/ was disfavored and realized as /xt/ (compare Ger. Luft ‘air’ and Kraft ‘power’ to Dutch lucht and kracht and English soft and after to Dutch zacht ‘soft’ and achter ‘after, behind’ where represenets [x]). It is thus impossible to say that postvocalic /xt/ is generally less favored than postvocalic /ft/. Rather, both /xt/ and /ft/ are “unstable” clusters and the one may turn into the other, depending on the local conditions within a language at a certain time: English developed in such a way that velar fricatives became disfavored (i.e. they became obsolete by replacement by other segments or by omission), whereas at a certain time, Dutch extended the distribution of velar fricatives (e.g. the Germanic voiced velar plosive is realized as /x/ in most dialects of Modern Dutch). Another consonantal innovation in Old English concerns allophonic variation among the fricatives: the singleton (or “short”) voiceless fricatives /f, θ, s/ became voiced word-medially in the context of other voiced phonemes, as in (4): (4)
SG
PL
wulf smiþ ( = /θ/) hu¯s
wulfas ( = /v/) ‘wolf’ smiþas ( = /ð/) ‘smith’ hu¯sa ( = /z/) ‘house’
After this allophonic rule had taken effect, no singleton voiceless fricatives were realized in word-medial position after a vowel. Old English intervocalic geminate consonants were degeminated later and this change gave rise to the phonemic opposition between short voiced and voiceless fricatives in word-medial positions. Thus, the addition of the phonological rule of degemination to the grammar generated segments (in this case singleton fricatives) that might have been the input of the phonological rule of intervocalic voicing. The fact that the degeminated fricatives did not undergo voicing indicates that the former allophonic rule ceased to have an effect. The result of these developments was that in word-medial positions, the opposition geminate versus singleton fricative was replaced by the opposition voiceless versus voiced fricative.
7 Linguistic Levels: Phonology The word-final contrast between voiced and voiceless fricatives developed through schwa-reduction of full vowels in unstressed position and eventual loss of final schwa between 1100 and 1300, i.e. the fricatives that were voiced by an allophonic rule before, were no longer followed by a vowel, but occurred in word-final position. The initial voiced fricatives /v/ and /z/ originated primarily from Norman-French borrowings in Middle English (e.g. valour ‘valor’, veel ‘veal’, and zele ‘zeal’). The result of the allophonic variation in Old English between voiceless and voiced fricatives combined with the loss of geminate consonants and the loss of word-final schwa and the introduction of /v/ and /z/ in initial position with borrowings from French resulted in a system where voiceless and voiced fricatives are distinctive.
2.3 Changes in phonotactics without effects on the consonant inventory In describing phonotactic restrictions, it is generally agreed in phonological theory that the concept of “sonority” plays a crucial role. Sievers (1881) and Jespersen (1904) introduced the Sonority Sequencing Principle to explain the fact that, within a syllable, the less sonorous segments are found at the periphery and the most sonorous sounds are found in the syllable peak. The idea that sonority is not an absolute property, but rather a relative one, gave rise to the notion of “strength hierarchies” (e.g. Lass 1970) and a “sonority scale”. In the sonority scale presented in (5), the degree of sonority of segments increases from left to right: (5)
Sonority scale for some segments p, t, k, b, d, ɡ f, s, ʃ, x m, n l r j, w i, u e, o a stops fricatives nasals liquids glides vowels of different height
In Old English, possible syllable-initial consonant clusters only needed to show a slight increase of sonority. In stressed syllables, initial clusters of voiceless /k/ and /x/ or voiced /ɡ/ followed by a nasal (/kn/, /ɡn/, /xn/) or clusters of the glide /w/ followed by a sound which was of the same degree of sonority (a rolled /r/) were as common as clusters with a steeper increase of sonority such as stop-liquid clusters. Lutz (1992) suggests that initial clusters with unfavorable phonotactics may result in different changes such as (i) the loss of the initial consonant of the cluster (e.g. ME wlispen → 14th/15th centuries lisp and ME fne¯sen → 14th/15th centuries neeze ‘sneeze’), (ii) the replacement of the initial consonant by a consonant that forms a more favorable consonant sequence (e.g. OE wlott → 14th century blot), and (iii) loss of words that start with such unfavorable clusters (e.g. OE wlank ‘proud’, wrabble ‘squirm’, and gnede ‘misery’ became obsolete in the 15th and 16th centuries). The syllable-initial clusters /xn/, /xl/, and /xr/ were lost within a short time span between the OE and the ME periods. The initial clusters /kn/, /ɡn/, and /wr/ were spelled and probably still pronounced as such when the writing system became more and more standardized after the introduction of Caxton’s printing press in the late 15th century. The
107
108
II Linguistic Levels reduction of /wr/ to /r/ probably took place relatively early (possibly starting in late Middle English). Shakespeare’s puns on knight–night, knot–not and wring–ring indicate that the initial sound in such clusters was either no longer pronounced during his time, or that there was variation between the pronunciation with and without the initial consonant. In the course of the 16th century, the phonotactics of the language gradually changed in such a way that sonority distances in syllable-initial clusters became larger: except for word-initial clusters with /s/ (spy, stop, sky, sneeze), the only permissible onset clusters after Shakespeare’s time are those in which a stop or fricative is followed by a liquid or a glide (as in the words dry, fly, cue, queen). In 17th-century educated English, the reduction of /kn/ and /ɡn/ to /n/ was completed and other dialectal varieties followed. In Present-day English, words that are borrowed from other languages that start with such impermissible clusters are modified by English speakers in such a way that both sounds of the cluster are realized. However, the ban on having syllable-initial /kn/ or /ɡn/ clusters still exists. In order to realize both members of the clusters, the strategy that speakers use today is to insert a vowel between the two consonants. Thus, the Hebrew word Knesset and the German name Knopf, for instance, are both pronounced with initial /kәn/ in Present-day English (see Green 1997: 25–28). Thus, the Present-day repair strategy for unfavorable phonotactics is vowel epenthesis.
2.4 Stability and instability in the vowel inventory and the effect on phonotactics Middle English diphthongs /iʊ/ (occurring in words like chew, due, and hue) and /eʊ/ (occurring in words like beauty, dew, and few) eventually collapsed under /juː/ (see Schlu¨ter and Jones, this volume). In the diphthong /iʊ/, the first part was reanalyzed as the glide /j/ and assigned to the onset, while /ʊ/ turned into /u:/ to compensate for the loss of vowel quantity in the late 16th century. In the diphthong /eʊ/, the first member was gradually raised to /i/ after the 16th century. In the 18th century, the element /i/ was reanalysed as a glide and the element /ʊ/ was tensed and lengthened, so that the original contrast between /iʊ/ and /eʊ/ was lost by the end of the 19th century in most dialects of English. The question is whether the original diphthongs palatalized the preceding consonant (as in djue and djew), or whether the original diphthongs are now realized as sequences of the palatal glide /j/ followed by the high back vowel /u:/. Consider in this respect that consonants with and without a following /j/ are not allophones of a phoneme, because – as the (6) illustrates – there are also instances of consonants followed by a high back vowel without the intervening palatal glide /j/: (6)
Syllable-initial sonorant a. music [mjuːzɪk] b. nude [njuːd] c. lurid [ljuːrɪd]
+ /j/ clusters in Present-day British English versus moose [muːs] versus noon [nuːn] versus lunatic [luːnətɪk]
Interestingly, even though the palatal glide [j] may follow [l] in word-initial position in Present-day British English, we never find that a cluster of a non-strident obstruent plus [l] is followed by [j]. Hence, there is no [j] between [l] and a following vowel in words like plumage, blue, clue, glue, but, according to Harris (1994: 61), some speakers have
7 Linguistic Levels: Phonology one in slew. This fact has important consequences for the analysis of English syllable structure. In particular, we might conclude from this observation that /j/ should not be analyzed as a secondary place of articulation. If it were, the explanation for the fact that a palatal glide may be present in single consonants, but not in some consonant clusters would be cumbersome. Instead, the phoneme /j/ is a consonantal segment which may occupy one position in the onset. Apart from /s/-initial clusters, the English onset may be filled by at most two positions and three-consonantal clusters such as */plj/, */blj/, */klj/ or */ɡlj/ would violate this restriction. Note that American English introduced a further restriction on initial clusters with /j/: it bans clusters in which a coronal consonant is followed by a palatal glide /j/ (as in the words tune, dune, suit, new). Since a palatal place of articulation is a sub-class of coronals, this ban may be interpreted as an OCP-effect (see Section 2.1).
3 The future of looking back in time on sound changes For Roman Jakobson and many phonologists after him, “phonology is the study of the properties of the sound systems which speakers must learn or internalize in order to use language for the purpose of communication” (Hyman 1975: 1). The word “systems” implies the notion of “stability” and suggests that the properties of sound systems are stable. However, as we saw in this chapter, little in phonology seems to be stable. For communication to be successful, speakers and hearers must share linguistic knowledge (e.g. knowledge of how sounds are used by the speaker). When we compare different stages of a language, we find that the properties of sounds that speakers use are not stable at all but nevertheless, communication between speakers and hearers functions well. How is this possible? Part of the answer lies in the fact that phonological properties of human languages are flexible and constrained only by what our vocal tract allows us to produce and by what our ears allow us to perceive. For instance, to differentiate voiceless stops from voiced ones, the following properties may vary: the length of the preceding vowel, the formants of a preceding and following vowel, the duration of the stop’s closure phase (i.e. closure duration), the presence/absence of vocal fold vibration during the closure phase (so-called “closure voicing”), the voicing lag or “VOT”, i.e. the time between the release of the closure phase and the onset of vocal fold vibration (with a large scale to choose from), burst intensity, burst duration, and more. In English, the major cues to distinguish laryngeal classes of stops are the length of the preceding vowel and the VOT values of the stops in question: voiced stops are preceded by a relatively long vowel duration and have a relatively short voicing lag whereas voiceless stops are preceded by a comparatively shorter vowel and have a longer voicing lag when initial in a word. Pohl and Grijzenhout (2010 and references therein) suggest that in German, a difference in voicing lag is the major cue to maintaining the contrast, whereas in Dutch closure voicing and the duration of the closure phase seem to matter most to identify a voiceless or voiced stop and in Swiss German closure duration alone contrasts two classes of stops. Thus, in the modern Germanic languages, each language favors a different cue or set of cues (and the speakers have different mental representations for the contrast in question). Children must “learn” to pay attention to the relevant cues in their language and then “internalize” that information – i.e. to build mental representations – to use it to convey meaning. Later generations may shift the attention to another cue or cues and this shift may
109
110
II Linguistic Levels be very small and hardly noticeable at first. When this shift is used by more speakers of a community, it may cause temporal or permanent change. For example, Dutch seems to shift from closure voicing and closure duration to differences in VOT as the major cue to maintain the distinction; the contrast between /p/ and /b/ remains, but its instantiation is different and it may lead to different mental representations for future generations. Thus, one source of change in the sound system is its inherent variability and flexibility which allows speakers to highlight relevant phonological properties and vary the irrelevant properties of sounds. Hearers too are able to ignore irrelevant phonetic detail (i.e. they abstract away from much information in the acoustic signal) and they pick up on just a few properties of the speech signal that make a distinction between two sounds. The system may change to enhance, create, or omit a contrast (e.g. a system with positive VOT distinctions of 10ms versus 80ms may become a system with negative VOT and positive VOT, or there may be room to add another VOT contrast or the contrast may vanish over time). Part of the answer why change occurs is thus: because the system is flexible, change is pre-programmed. The next question is when does change occur? The Neogrammarians advocated the view that change is “spontaneous” and others also find an answer in sociological factors (e.g. peer group pressure, the need to be different from your neighbor, the need to borrow words from your neighbor). Human beings live in groups. In order for the group to be successful, it has either to blend in with other groups (so that a language contact situation may result in language change), or to compete with other groups (so that variations emerge within one dialect or language which may eventually lead to language split) or – when the group becomes too big – to form subgroups with their own identity (and, hence, their own language variety). This is a continuing or “dynamic” process. Languages will always change and there is enough flexibility in the speech signal to allow them to do just that. The question that many people raise is why some languages seem to resist change or change relatively slowly, whereas others seem to change more rapidly. To illustrate the point, we would here like to compare the histories of Icelandic and English. Icelandic is spoken in a relatively isolated part of the world where it did not come in contact with many other languages (one possible source of language change). Moreover, as Kristja´n ´ rnason (p.c.) once pointed out to me, there has been a long tradition of a conscious A effort to maintain the language and not to allow innovations or to “imitate” the Danes and the Germans. Dialectal variations that emerged in smaller communities in Iceland were stigmatized and gave way to the standard dialects spoken by the majority. Moreover, the population is small enough for language planning committees to be able to impose their recommendations and social pressure has a major effect on the relative stability of the language. In contrast, ever since the Anglo-Saxons landed on the British Isles, the English language has been exposed to foreign influences and this has had a major effect on the relative instability of some of its properties. There have been many attempts to preserve or “improve” the language, but such attempts apparently have been less successful than in Iceland (see Percy, Chapter 63). As late as at the turn of the 20th century, phoneticians like Daniel Jones attempted to develop a theory of sound systems that would help second language learners to acquire the accurate pronunciation of words in a foreign language (Jones 1917). Moreover, phonetics as the study of sound systems was considered to be useful in helping to improve the pronunciation of native speakers. George Bernard Shaw certainly contributed to this view in
7 Linguistic Levels: Phonology 1912 when he made Professor Higgins teach Eliza Doolittle the “proper” pronunciation of words in his popular play Pygmalion. Such attempts have proven to be futile: there is no proper pronunciation of English and the many speakers of the language will never employ exactly the same pronunciation of all sounds. English has always had many dialects: some are conservative and have archaic features, whereas others are more progressive and are open to innovations. Both types of dialects have considerable numbers of speakers for whom it is advantageous (from a sociolinguistic perspective) to use the dialectal variety. Also, since the 16th century, the language has expanded to new territories and the distance between the British Isles may have lead to independent developments between varieties of English spoken throughout the world. There are thus many reasons why English has changed so much over the centuries and why there is so much diversity. The future of looking back on English sound change lies in the fact that English is a global language with a rapidly growing number of speakers and for this reason alone, interest in its ever changing sound system will not vanish. Even though this handbook is voluminous, it will still be incomplete in that not all changes which have occurred in the English language will be illustrated and discussed. Improved techniques, the availability of more extensive corpora, etc. will in the near future certainly result in extending our knowledge of which aspects triggered which kind of change in the history of the English language. Linguists like Aditi Lahiri and Frans Plank (p.c.), however, hold that it is actually more surprising that some aspects of a language are stable and do not change (or change relatively slowly or only within a relatively small area), i.e. the future of looking back on linguistic change also lies in the interest in which properties of language are unstable and highly variable compared to properties that are stable or invariable in the sense that they resist change or change slowly. This is a relatively underexplored field.
4 Summary Phonology is the study of the properties of sounds which speakers use to convey meaning. As pointed out in Section 1, a primary goal of structural and generative phonology is to investigate the function of speech sounds, i.e. to establish and compare phoneme inventories of languages. In this chapter, we have seen examples of phonemes that changed their shape (e.g. voiceless stops becoming fricatives according to Grimm’s Law), phonemes that were added to the English inventory (e.g. alveopalatal fricatives), and phonemes that became obsolete (e.g. /kw/, /x/). We have also looked into the behavior of speech sounds: some allophonic rules (e.g. voicing of fricatives in the context of other voiced segments) emerged at some stage and disappeared again later. Section 2.2 discussed two examples of allophonic variants that developed into phonemes (alveopalatal affricates and voiced fricatives, respectively). Another goal of phonology is to account for the distribution of speech sounds. For English, we have seen that changes in phonotactics may be temporal (e.g. the ban on /sk/) and that the effect of a phonotactic restriction may give rise to a new phoneme (e.g. /ʃ/). Some phonotactic restrictions may be added relatively late in the history of a language (e.g. the ban against /kn/ clusters) and may change their effect (from deletion of a segment in the cluster to insertion of a vowel to break up the cluster). Sometimes, a change in the phoneme inventory may result in the addition of new
111
112
II Linguistic Levels phonotactics (e.g. changes from the diphthongs /iʊ/ and /eʊ/ to the monophthong /uː/ preceded by /j/ which can form a complex onset with any single consonant). In many cases, a merger of two consonants into one may result in a new phoneme for which the distinctive features are already part of the system (/sk/ → /ʃ/), allophonic rules and borrowings may also create new phonemes (e.g. voiced fricatives) and a change from a complex phoneme into a simplex one may result in new phonotactics.
5 References Aitchison, Jean. 2001 [1981]. Language Change: Progress or Decay? 3rd edn. Cambridge: Cambridge University Press. Baudouin de Courtenay, Jan Niecisław. 1895. Versuch einer Theorie phonetischer Alternationen. Strasbourg: Tru¨bner. Goldsmith, John A. 1995. Phonological theory. In: John A. Goldsmith (ed.), The Handbook of Phonological Theory, 1–23. Cambridge, MA/Oxford: Blackwell. Green, Anthony Dubach. 1997. The Prosodic Structure of Irish, Scots Gaelic, and Manx. Ph.D. dissertation, Cornell University. Harbert, Wayne. 2007. The Germanic Languages. Cambridge: Cambridge University Press. Harris, John. 1994. English Sound Structure. Cambridge, MA: Blackwell. Hyman, Larry M. 1975. Phonology: Theory and Analysis. New York: Holt, Rinehart and Winston. Iverson, Gregory K. and Joseph C. Salmons. 1995. Aspiration and laryngeal representation in Germanic. Phonology 12: 369–396. Jakobson, Roman. 1939. Observations sur la classement phonologique des consonnes. Proceedings of the 3rd International Congress of Phonetic Sciences: 34–41. Jakobson, Roman, Gunnar Fant, and Morris Halle. 1951. Preliminaries to Speech Analysis. Cambridge, MA: MIT Press. Jakobson, Roman and Morris Halle. 1956. Fundamentals of Language. The Hague: Mouton. Jakobson, Roman. 1971 [1962]. Retrospect. In: Selected Writings I: Phonological Studies, 631–658. The Hague: Mouton. Jespersen, Otto. 1904. Lehrbuch der Phonetik. Leipzig: Teubner. Jones, Daniel. 1917. An English Pronouncing Dictionary. London: Dent. Jones, Daniel. 1967. The Phoneme; its Nature and Use. 3rd edn. Cambridge: W. Heffer & Sons Ltd. Kiparsky, Paul. 2003. The phonological basis of sound change. In: Brian D. Joseph and Richard D. Janda (eds.), The Handbook of Historical Linguistics, 313–343. Malden, MA: Blackwell. Lass, Roger. 1970. Boundaries as obstruents: Old English voicing assimilation and universal strength hierarchies. Journal of Linguistics 7: 15–30. Lass, Roger. 1984. Phonology: An Introduction to Basic Concepts. Cambridge: Cambridge University Press. Lutz, Angelika. 1992. Lexical and morphological consequences of phonotactic change in the history of English. In: Matti Rissanen, Ossi Ihalainen, Tertta Nevalainen, and Irma Taavitsainen (eds.), History of Englishes, 156–166. Berlin/New York: Mouton de Gruyter. McCarthy, John J. 1993. A case of surface constraint violation. Canadian Journal of Linguistics 38: 169–195. McCarthy, John J. and Alan S. Prince. 1993a. Generalized Alignment. In: Geert Booij and Jaap van Marle (eds.), Yearbook of Morphology 1993, 79–153. Dordrecht: Kluwer. McCarthy, John J. and Alan S. Prince. 1993b. Prosodic Morphology I: Constraint Interaction and satisfaction. Ms., University of Massachusetts, Amherst and Rutgers University. McCarthy, John J. and Alan S. Prince. 1995. Faithfulness and reduplicative identity. In: Jill N. Beckman, Suzanne Urbanczyk, and Laura W. Dickey (eds.), University of Massachusetts Occasional Papers 18: 249–384.
8 Linguistic Levels: Prosody
113
Pohl, Muna and Janet Grijzenhout. 2010. Phrase-medial bilabial stops in three West Germanic languages. Linguistische Berichte 222: 141–167. Prince, Alan and Paul Smolensky. 1993. Optimality Theory: Constraint interaction in Generative Grammar. Ms., Technical Report no. 2, Rutgers University Center for Cognitive Science, Cambridge, MA, MIT Press. Sapir, Edward. 1921. Language: An Introduction to the Study of Speech. New York: Harcourt, Brace; downloaded from http://www.bartleby.com/186/. de Saussure, Ferdinand. 1916. Course de linguistique ge´ne´rale. Paris: Payot. Sievers, Eduard. 1881. Grundzu¨ge der Phonetik. Zur Einfu¨hrung in das Studium der Lautlehre der Indogermanischen Sprachen. Leipzig: Von Breitkopf & Ha¨rtel. Waugh, Linda R. 1976. Roman Jakobson’s Science of Language. Lisse: The Peter de Ridder Press. Wunderlich, Dieter. 2008. Spekulationen zum Anfang von Sprache. Zeitschrift fu¨r Sprachwissenschaft 27: 229–265.
Janet Grijzenhout, Konstanz (Germany)
8 Linguistic Levels: Prosody 1 2 3 4 5 6 7
Definition of terms Syllable structure and syllable weight Historical sources of information for prosodic reconstruction Old English meter and prosody Middle English meter and prosody Post-Middle English prosodic innovations References
Abstract The chapter traces the development of word and phrasal stress from Old to Present-day English. Section 1 and Section 2 define the terms needed to describe the prosodic patterns of speech and address the notions of syllable structure and syllable weight. Section 3 surveys the methodological bases for prosodic reconstruction, focusing specifically on the interplay between meter and language in the recovery of rhythmic patterns in speech. Old English meter and prosody are covered in Section 4, where the basic principles of Old English alliterative versification provide the foundation for reconstructing word and phrasal stress. Middle English meter and prosody are covered in Section 5, again with specific references to metrical form, word stress, and phrasal stress. The section includes a discussion of the effect of lexical borrowing from French and Latin on the prosody of English. Section 6 is devoted to the major prosodic changes in English during and after the Renaissance.
1 Definition of terms The term “prosody”, as used in this chapter, refers to the properties and the organization of syllables into words, phrases, and sentences in speech. Outside of linguistics, the term prosody can also be used with reference to the study of verse and its properties; Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 113–128
114
II Linguistic Levels for the conventionalized rhythmic structures of verse we reserve the term “meter”. The prosodic properties of speech are “suprasegmental”: their domain is larger than individual speech sounds, which are organized into higher-level units that are independently pronounceable, namely “syllables”. The ability to divide an utterance into syllables is part of the intuitive knowledge that speakers have of their language. Very importantly, syllables are the carriers of “stress”, the contrastive intensity that marks some syllables as more or less prominent. Phonetically, stress is associated with the use of a greater amount of respiratory energy on a syllable, increased tension of the vocal folds, and loudness. In terms of metrical structure, the prominent position is called an “ictus” (S); ictic positions are usually, but not always, filled by stressed syllables, while “non-ictic” (W) positions attract unstressed syllables. Stress is binary in the sense that syllables are either stressed or unstressed. Further, a stress may range from a full primary/main stress, here marked with ´ (acute), to various levels of non-primary stress, here marked with ` (grave). Although informally we speak of “stressed” and “unstressed” vowels, and we place the stress marks over the vowels for typographic convenience, it is important to bear in mind that stress is a property of the entire syllable.
2 Syllable structure and syllable weight The syllable is the smallest pronounceable prosodic unit, but it is also structurally complex in that it is further decomposable. At the core of the syllable is its “nucleus” or “peak”, the segment of highest sonority in the string. Every syllable has to have one and only one nucleus, usually a vowel or a diphthong, but sometimes also a syllabic sonorant /r̩ , l̩, m̩ , n̩ /. Consonants or consonant clusters to the left of the nucleus constitute the syllable “onset”, and the consonants following the nucleus make up the “coda”. The onset and the coda are not obligatory elements of the syllable. Universally, a filled onset is preferred to a filled coda. A coda consonant can contribute to the “weight” of a syllable, whereas an onset is commonly considered weight-neutral. The division of a string of sounds into syllables follows the “Maximal Onset Principle”. According to that principle, a single consonant between two vowels fills the onset of the syllable to the right (syllable divisions are marked with a period): rea.son, e.ne. my, de.hu.mi.di.fy. A two-consonant cluster is either divided or not, depending on whether the resulting onset is also a possible word-initial cluster: com.post, prag.ma. tic, fic.tion, but hi.sto.ry, pa.tri.ot. Three consonants between vowels are split again depending on the nature of the resulting cluster: emp.ty, friend.ly, coun.try, um.bre.lla, a. strin.gent, o.sprey. The Maximal Onset Principle does not apply across prosodically independent words, so boil eggs is not *boi.leggs. Syllable weight is a prosodic property tightly associated with stress: universally, heavy syllables attract stress and syllables that carry stress are likely to become heavy. In English a “heavy syllable” is any syllable whose peak is a long vowel or a diphthong: see.saw, pay.ee, or any syllable that ends in a consonant: com.pul.sion, prac.tice. A “light syllable” has a short vowel in the peak and no coda: A.me.ri.ca, re.pli.ca. In practice, in Present-day English all syllables except those ending in /ɪ, ɛ, æ, ʊ, ʌ, ə/ are heavy. Monosyllabic major class words (clue, club, day, fry, three, wet) cannot have a light syllable; it follows that /*clɛ, *frɪ, *sʊ/ would not be possible
8 Linguistic Levels: Prosody English words, while /clɛs, frɪn, sʊg/ are possible words which are accidental gaps in our vocabulary.
3 Historical sources of information for prosodic reconstruction Reconstructing the prosodic properties of the earlier stages of English is a complex task. The types of segmental changes that occur in stressed and unstressed syllables are very dissimilar. Vowel lengthening, vowel shifting, and gemination typically occur in stressed syllables, while vowel reduction and loss and consonant lenition are expected in unstressed syllables. If we find textual evidence of such processes, we can make prosodic inferences: the progressive reduction and loss of the prefix ge- (OE geriden > ME iriden > PDE ridden ‘ridden’) is good evidence that ge- was unstressed in Old English and Middle English. Similarly, Middle English spellings luved, luvd for earlier luvede ‘loved’ indicate reliably that the form was initially stressed. Our most direct source of information about the prosodic structure of earlier English, however, comes from the way in which the forms of speech are matched to the structural positions in verse. The greatest challenge for the use of verse as the primary evidence for prosodic reconstruction is circularity: since there are no records of instructions on what is permitted in early versification, we rely on templates extrapolated from the surviving poetic corpus. Our understanding of how the metrical templates worked is thus founded on a web of typological inferences about language and meter with no possibility of direct verification. The way we avoid ignotum per ignotius, explaining ‘the unknown by means of the more unknown’ is by applying testable quantitative and typological criteria to the formulation of the rules of meter and the reconstruction of prosodic patterns. The statistical data on some features, e.g., in 26,088 verses of OE poetry, only 36, or 0.001%, lack alliteration (Hutcheson 1995: 169), justify reliance on alliteration as the binding principle in the alliterative long line, here marked in boldface. Moreover, testably unstressed syllables, such as inflectional syllables, never alliterate, which makes the co-occurrence of stress and alliteration a solid source of prosodic reconstruction. No matter what theory of Old English meter one adopts, there can be no doubt that in Beowulf (henceforth Beo) 102: wæs se grı´mma gæst / Gre´ndel haten ‘was the grim ghost / Grendel called’, the words grı´mma and Gre´ndel are initially stressed. Typologically too, all Germanic languages, including Present-day English, stress native unprefixed words on the first syllable; we can safely project that back to Old English and posit ´ ðen ‘heathen’, so´þe ‘truly’. root-initial stress on cy´ning ‘king’, de´maþ ‘they judge’, hæ The alignment of the main stress with the left edge of a simplex word in early English is known as the “Germanic Stress Rule” (GSR). Statistical and typological grounding of prosody-meter correspondences is our best recourse in spite of some inherent uncertainties. The historical poetic corpus presents cases where deviations from an established norm may be interpreted as deliberate creative choices. A poet may force an unstressed syllable into an ictic position to fit the expectations of the template: thus Chaucer rhymes felawe : awe, biddyng : thing. This convention of versification is of no use to us in trying to reconstruct the prosodic contour of Germanic felawe or biddyng in speech – the words were always initially stressed. On the other hand, Chaucerian rhymes such as honour : flour, servise : wyse have, all too freely, been taken as evidence for non-initial stress on the Romance borrowings honour
115
116
II Linguistic Levels and servise. Such evidence has to be evaluated carefully and compared to the evidence of the placement of such words in line-medial position. Similarly, placing re´ady, u´nder, ma´keth at the left edge of an iambic (W S) line is a metrical inversion which breaks the monotony of repeated identical structures, but it tells us nothing new or special about the prosody of these native words. However, placing loans such as citees, justice, poynaunt line-initially is open to both WS and SW meterical scansion and can be considered good evidence that such words maintain their Romance stress contour. We will return to these metrical issues in Section 5.1. For now we just note that decisions on the prosodic history of loanwords will have to be based on fine-grained and comprehensive coverage of the placement of individual items in the verse.
4 Old English meter and prosody Germanic and Old English versification is notoriously difficult to model. Although new theories of Old English meter continue to appear, most recently in Getty (2002) and Bredehoft (2005), no new approach rivals the descriptive adequacy and scholarly acceptance of the observations and patterns in Sievers (1893); see Stockwell and Minkova (1997), Minkova (2008a).
4.1 Basic principles of Old English alliterative verse Sievers’s hypothesis about the metrical structure of Old English verse rests on the following configurations: – a line consists of two verses, the “on-verse” (“a-verse”) and the “off-verse”, (“b-verse”), linked by alliteration; – each verse contains two feet and at least four positions; – each foot contains an ictus (S), also known as a “lift”, and at least one non-ictic position (W), also known as a “dip”. This allows us to represent the structure of the line as in Figure 8.1, where the numbers at the bottom stand for positions: LONG LINE
ON-VERSE/a-VERSE
FOOT
1 ne
OFF-VERSE/ b-VERSE
FOOT
2 leof
‘not friend
3 ne
FOOT
4 lað
nor enemy
1 be
FOOT
2 lean
dissuade
3 mih
4 te
could’ (Beo 511)
Figure 8.1: The structure of the Old English verse line
8 Linguistic Levels: Prosody The binary representation in Figure 8.1 is an abstraction based on the minimal line structure in terms of syllable count. The prominence relations are unspecified; within the feet, lifts and dips can appear in either order. Each position is ideally filled by a single syllable, and an S position must be filled by at least one syllable. Unlike the familiar notion of classical metrical feet, positions and feet in Old English verse may be of uneven size, due mainly to the expandability of the non-final weak positions in each ´ re ‘thanes are united’ (Beo 1230a) the template S verse, thus in þe´gnas syndon geþwæ W S W has the first W position filled by four unstressed syllables: -nas syn.don ge-. The one-to-one correspondence between a syllable and a position may be disregarded for S-positions under special metrical conditions; this is known as “resolution”. Resolution is a metrical equivalence: one and only one heavy syllable can fill a lift, but a light syllable and any other syllable may jointly fill a lift to avoid an unacceptable metrical violation, such as an expanded dip at the right edge of the verse. Thus in the S W S W verse re´ceda under ro´derum ‘of halls under heavens’ (Beo 310a), the syllables ro´. de- are metrically subsumed under the second S position to avoid the unacceptable matching of the last W to -de.rum. The conventions of alliteration which help us separate relevant from irrelevant metrical information are: – in the on-verse both S positions may alliterate. – in the off-verse only the first S position is allowed to alliterate. Nearly all verses are complete syntactic units. The smallest linguistic units that occupy a verse are compounds, e.g. þeodcyninga ‘of tribe-kings’ (Beo 2a), wilgesiþas ‘willing companions’ (Beo 23a), landgemyrcu ‘shore-boundaries’ (Beo 209b). Most often, however, a verse is coextensive with a clause or a syntactic phrase: Hi hine þa ætberon / to brimes faroðe ‘they him then carried / to the sea’s current’ (Beo 28). An intriguing convention, not fully understood, describes the hierarchy of syntactic elements within the verse with respect to alliteration. In a verse where the S-positions are filled by a noun and a verb, the noun will consistently be strong, whether it is an NP-VP string as in Him ða Scyld gewat ‘Then Scyld departed’ (Beo 26a), or a VP-NP: Gebad wintra worn ‘Lived to see winters many’ (Beo 264a). This alliterative regularity is known as Sievers’s “Rule of Precedence” (Sievers 1893: Sections 22–29); it states that if an inflected verb precedes a noun it does not have to alliterate, that it must not alliterate if the noun does not alliterate too, and that a non-alliterating noun can never be followed by an alliterating finite verb. The rule does not exclude double alliteration: þenden wordum weold ‘when with words ruled’ (Beo 30a), geafon on garsecg ‘gave in ocean’ (Beo 49a), so projecting the Rule of Precedence on to the prosody of Old English speech is not always straightforward. The significance of alliteration in the reconstruction of phrasal and utterance prosodic contours will be discussed further in Section 4.3.
4.2 Old English word stress As noted in Section 3, Old English word stress falls on the first stressed syllable of word roots. The acoustic prominence of stress is thus, unsurprisingly, an important and consistent morphological boundary signal. All root-initial syllables are stressed. The weight of the root-initial syllable is irrelevant; both heavy and light syllables can be stressed: drı¯´.fan
117
118
II Linguistic Levels ‘drive’, fu´l.tum ‘help’, me´n.gan ‘mix’, so´¯ .na ‘soon’ (heavy), and cy´.ning ‘king’, ga´.fol ‘tax’ me´.du ‘mead’, sca´.mu, ‘shame’ (light). A very important difference between Old English and Present-day English is the stability of stress on the first root syllable in a derivational set: while suffix-induced stress-shifts in Present-day English can leave root-initial syllables completely stressless: chronic–chrono´logy, ´ıdiot–idio´tic, so´lid–solı´dity, Old English word roots are always marked by the presence of stress: ge´ogoð ‘youth’ hla´ford ‘lord’
ge´ogoðhad ‘youth-hood’ hla´fordscipe ‘lordship’
wo´ruld ‘world’ wu´ldor ‘glory’
wo´ruldlic ‘worldly’ wu´ldorfull ‘glorious’
The addition of suffixes in Old English never affects the primary prominence. The suffixes themselves can bear some degree of non-primary stress because they can be ictic, but they are automatically excluded from the positions of obligatory alliteration, the first ictic positions in each verse. Inflectional suffixes are always unstressed, while derivational suffixes exhibit complex behavior in the verse and it is likely that their prosodic realization in speech was gradient, ranging from non-primary stress to absence of stress. The variability is attested both synchronically and diachronically. The position of the suffix with regard to the word boundary is of relevance, and so is vowel quality and quantity. When inflected, heavy suffixes with long non-high vowels (-le¯as- ‘-less’, -do¯m- ‘-dom’, -fæst- ‘-fast’ -ha¯d- ‘-hood’) are regularly scanned as lifts, e.g. wı´sdo`me heold ‘with wisdom ruled’ (Beo 1959b), of cı´ldha`de ‘from childhood’ (Elene 914a), but uninflected -do¯m ‘-dom’, -fæst ‘-fast’ are not ictic: word ond wı´sdom ‘word and wisdom’ (Andreas 569a), wı´sfæst wo´rdum ‘wise with words’ (Beo 626a). The placement of the word linearly in the verse is also significant: the suffixes -sum ‘-some’, -scipe ‘-ship’, -ian ‘-en’ (V) occupy ictic positions only in the coda of the verse; for full coverage see Fulk (1992: 197–216). Further indeterminacies arise from the difficulty of assigning suffixal status to morphological units which are also attested as independent words: do¯m, fæst, full, ha¯d, le¯as are separate lexical entries, and their semantic autonomy may be related to the preservation of stress. Additionally, as demonstrated in Minkova and Stockwell (2005), the full prosodic history of native suffixes has to refer to rhythmic factors linked to the types and frequency of derived words in the lexicon. Thus the equally productive OE suffixes -ha¯d and -do¯m would be expected to emerge either both with a full vowel, or both with a reduced vowel in Early Modern English. However, in Middle English close to 70% of the -dom derivatives followed a monosyllabic root (earldom, freedom, kingdom, wisdom), where stress-clash avoidance resulted in de-stressing of the suffix to [-dəm/-dm̩ ], while during the same period 73% of -hood derivatives had a disyllabic stem (bishophood, maidenhood, womanhood), allowing the preservation of secondary stress on the affix and raising of the long vowel to [uː] prior to 17th-century shortening to [-hʊd]. In summary, all factors identified above – syllable weight, vowel quality or quantity, semantic independence, and rhythmically induced changes – must be considered in the account of Old English suffixal stress. Derivational affixes often have their diachronic roots in independent words. Within the larger family of affixes, suffixes are cross-linguistically more likely to lose their
8 Linguistic Levels: Prosody
119
independent word status than prefixes, and therefore one would expect more word-like behavior from prefixes. Identifying the exact range of prefixes in Old English is a widely recognized problem, precisely because outside of the invariably bound forms: æf-, and-, be-, ed-, fær-, for-, ge-, mis-, etc., there is no clear-cut divide between prefixes such as ofer-, on-, wiþ-, ymb- and words. Moreover, the metrical treatment of both bound and free prefixal forms may differ for nouns and adjectives, where main stress aligns with the left edge of the whole word, leaving the root with secondary stress, and verbs and adverbs, where the main stress is kept on the root: compare swylce o´ncy`þðe ‘such grief’ (Beo 830a) to he onfe´ng hraþe ‘he seized quickly’ (Beo 748b) As argued in Minkova (2008b), both syllable weight and the grammatical nature of the base are determiners of stress in OE prefixation. Light prefixes behave like clitics; they do not form independent prosodic words and are consistently unstressed, while prefixes capable of forming independent prosodic words get stressed in accord with the word class of the derivative. NO
V, ADV Morphological Representation
Prefix ≠ PRWD
NO
Prefix = PRWD
YES
N, ADJ
Figure 8.2: Prefixal stress in Old English (adapted from Minkova 2008b: 36)
The principle of root-initial stress persists in compounding, where roots get their first syllables stressed as if they were independent words. Within the larger domain of compounds the stress to the left is primary, marking off the left boundary of the entire word, while compound-internal stresses are secondary. In the verse, the obligatory alliteration is consistently placed on the first stressed syllable onset, e.g. ofer hro´nra`de ‘over whaleroad’ (Beo 10a), wo´rolda`re forgeaf ‘worldly honor gave’ (Beo 17b). The second stressed syllable may alliterate only if the first stress alliterates too: wið þe´odþre`aum ‘against people’s calamity’ (Beo 178a), he´ardhı`cgende ‘hard-minded’ (Beo 394a). Such selfalliterating compounds are restricted to the on-verse by definition, since alliteration is prohibited from the second ictus in the off-verse. This restriction does not extend to affixal elements, thus la´ðlı`ce ‘hatefully’ is found at the right edge of the off-verse. The inference is clear: in þe´odþre`a ‘people-calamity’, he´ardhı`cgend ‘hard-minded’, both roots retain their semantic independence and strong prosodic prominence. Such forms present an analytical problem: they are interpretable both as compounds and as freely formed syntactic phrases. Another difficulty comes from the fact that many of the self-alliterating compounds in the Old English corpus are hapax legomena, single-instance forms: be´arngeby`rdo ‘child-bearing’, e´all-ı`ren ‘all of iron’, fe´n-fre`oðo ‘marsh refuge’, gry´re-ge`atwe ‘terrifying armor’, gry´re-gı`est ‘terrible visitor’, he´ardhı`cgend ‘hard-minded, he´oro-ho`cyht ‘savagely hooked’, hı´lde-hlæ` mm ‘battle crash’, swa´t-swa`ðu ‘bloody track’, sy´n-snæ` d ‘huge cut’, þe´odþre`a ‘people-calamity’ are some examples of such unique forms in Beowulf. The status of these constructions is an area deserving further inquiry; cf. Giegerich (2009) who shows that end-stress on noun-noun compounds in Present-day English: steel bridge, apple pie, Madison Avenue, may reflect the syntactic provenance of incompletely
120
II Linguistic Levels lexicalized forms, and that nominals of the form attribute-head can be both lexical and syntactic.
4.3 Old English phrasal stress In connected speech words are grouped together in larger prosodic constituents: clitic and phrasal groups. Clitic groups are made up of a fully stressed head-word and a clitic, an unstressed function word such as an article, a preposition, a conjunction, or a pronoun: the bo´ok, at scho´ol, etc., are clitic groups. Such groups behave in the same way in Old English: se rı´ca ‘the ruler’, on be´arme ‘on bosom’. The stressed words in a sentence are syntactically organized into noun-, verb-, adverb- and adjective phrases, coordinate phrases, and clauses. In Present-day English such syntactic units are right-prominent; i.e., the highest prominence is on the rightmost stressed syllable, while other stresses are secondary: ca`reful respo´nse, drı`ve ca´refully, ve`ry ca´reful, qu`ick and ca´reful, Be`n ca´res. Recovering the corresponding prosodic features of Old English from the existing textual records is challenging, and therefore the issue of phrasal prosody is underresearched and controversial. We know with certainty that the poets treated finite verbs differently from nouns; see Section 4.1 for Sievers’s Rule of Precedence. ´ l ala`mp ‘until time came’ (Beo 622b)) are Clause-final intransitive verbs (oþþæt sæ metrically weaker, but this convention may not match speech prosody. Clause-initial finite verbs may be skipped by the alliteration, e.g. Co`m þa to re´cede ‘came then to building’ (Beo 720a), Forge`af þa Be´owulfe ‘gave then to Beowulf’ (Beo 1020a), although the verbs also occupy ictic positions. Like the second elements of compounds, finite verbs may alliterate only if the other stressed word in the verse alliterates: we`ox under wo´lcnum ‘waxed under the clouds’ (Beo 8a), … wo´rdum we`old ‘… with words ruled’ (Beo 30a). Throughout the modern Continental West Germanic languages and in older Germanic, complements are stronger than their verbs, irrespective of the linear order. This typological comparison and the consistency with which complement-verb prosodic relations are observed in the verse – the complement always alliterates – is a good argument for projecting this prosodic contour to Old English. Prominence in noun and adjective phrases and coordinate phrases is not directly recoverable from the verse. In this area the rules of alliteration may be more of a handicap than help. The frequent assumption that obligatory alliteration on the first word in such phrases (lange hwile ‘a long while’ (Beo 16a), … hond ond rond ‘… hand and shield’ (Beo 656a)) translates directly into left prominence in the prosody is unfounded. The linear alliterative arrangement is a purely metrical convention, as can be seen from the freedom with which the poet switches components to fit the scheme in the line: Geata dryhten ‘lord of the Geats’ (Beo 2561b) vs. dryhten Geata (Beo 2901a); madma fela ‘of treasures many’ (Beo 36a) vs. fela missera ‘many of half-years’ (Beo 153b), manig oðerne ‘many other (men)’ (Beo 1860b) vs. æþeling manig ‘hero many’ (Beo 1112b). Some other facts also prompt skepticism about the link between alliteration and linguistic prominence: the default contour (no special focus) for noun phrases and coordinate phrases in the modern Germanic languages is right-prominent; for German see Selkirk (1984: 225–230). Right-hand prominence is attested also in copulative combinations of the type Anglo-Sa´xon, Native Cana´dian; they also typically align with
8 Linguistic Levels: Prosody syntactically coordinated phrases. The density of double alliteration in on-verses coextensive with noun + prepositional phrase (bat under beorge ‘boat under cliff’ [Beo 211a]) and in conjoined phrases (word ond wı´sdom ‘word and wisdom’ [Andreas 569a]), exceeds by far the overall 47% ratio of double alliteration in the on-verse, as reported in Hutcheson (1995: 112). This asymmetrical distribution precludes a linguistic bias towards left-prominence, but does not rule our equal or right-hand prominence. The absence of double alliteration in the off-verse can only be metrically determined; see Russom (1987: 114), Hutcheson (1995: 271). The most economical account that does not require a historical shift, therefore, is that the right-prominent prosodic contour of phrasal stress has been in the language since Old English times.
5 Middle English meter and prosody The Norman Conquest of 1066 coincides roughly with the abandonment of the structural principles of Classical Old English alliterative versification. The last surviving pieces of alliterative poetry that conform to the norms outlined in Section 4.1 are two short poems: Durham, c.1100, and The Grave, c.1150. Early Middle English compositions such as The Proverbs of Alfred, The Worcester Fragments of the Soul’s Address to the Body, The Bestiary, and Lagamon’s Brut, are “hybrid” compositions, mixing rhyme, alliteration and syllable-counting in often erratic patterns. Being grounded in the prosodic pattern of stress on the first root syllable, alliteration as a cohesive device survived, and a significant portion of the literary activity in the 14th century was channeled into the reinvention and composition of alliterative verse, culminating in masterpieces like Sir Gawain and the Green Knight, Piers Plowman, and The Alliterative Morte Arthure. At the same time new modes of versification based on rhyme, stress alternation, and syllable counting were gaining popularity. The relative rigidity of the new forms provides a solid basis for reconstructing the prosodic properties of Middle English.
5.1 Middle English metrical innovations: isosyllabicity, rhyme, iambic feet Verses of equal numbers of syllables – “isosyllabic verses” – are not uncommon in Old English poetry, but the recurrence was not structurally regulated; a verse could have from a minimum of four to fourteen syllables. Isosyllabism is an imported metrical feature in Middle English. Schemes based on the iteration of isosyllabic lines – the octosyllabic line, the septenarius, and, with Chaucer, the decasyllabic iambic pentameter – are at the core of Middle English verse composition. All of these forms allow an unstressed syllable after the last ictus; such “extrametrical” syllables are outside the metrical template and their presence or absence does not affect the isosyllabicity of the line. The lines were often linked in couplets or larger groups by end-rhymes. Rhyming did appear occasionally as an ornamentation in Old English verse, but the influence of Anglo-Norman made it the verse-line marker of choice. The third component of the new type of versification is the “iamb”, a binary sequence of a weak and a strong position (W S). Iambic feet could occur in Old English verse as a subset of a larger right-strong metrical type: the first three feet in ne le´of ne la´ð / bele´an mihte ‘not friend nor enemy / dissuade could’ (Beo 511; see Figure 8.1),
121
122
II Linguistic Levels happen to be iambs. In Middle English isosyllabic verse, however, the iamb became the dominant metrical foot. The earliest post-Conquest long non-alliterative compositions, The Owl and the Nightingale and The Ormulum, both late 12th century, are strictly iambic. Chaucer’s poetic works are also iambic. (All Middle English verse examples in this chapter are from Chaucer; abbreviations are from The Riverside Chaucer [Benson ed. 1987: 779].) The reconstruction of stress based on the new type of versification is most reliable line-medially. The interplay between prosody and meter at the two edges of the line is complicated by specific properties of the first and the last foot. The left edge of the line is rhythmically malleable, so that the expected W S / W S metrical cadence of the first two feet may be filled by: – a prosodic /s w w s/. An inverted foot S W followed by a regular W S foot is known as ´ n.der his belt … (GP 105). Triples may a “triple”: Tho´n.ked be Go´d … (WBT 5), U appear elsewhere in the line, but the probability of a triple decreases sharply from left to right. – a prosodic /s w s w/, resulting in trochaic inversion in both feet (Spo´ones and sto´oles and … (WBT 288), Ha´rdy he wa´s and … (GP 405). Occasionally whole lines can be trochaic: Ble´ssinge ha´lles, cha´mbres, kı´chenes, bo´ures (WBT 869). – by /Ø s w s/, where the W of the first foot is unfilled and the line is headless: Twe´nty bo´okes, clad in blak or reed (GP 294), Swe´re and ly´en, as a womman kan (WBT 228). The strong position in the rightmost foot of the line, where the rhyme is located, is metrically demanding in that it enforces prominence on the syllable filling that position. This is a verse convention, possibly observed in recitation, but it does not carry over into the prosody of speech. In Middle English rhyming practice, some suffixes appear to acquire metrically-induced secondary stress: bo´ldely`, dro´nkene`sse. The metrical strictness of the last strong position is such that it can even invert the prosodic contour of a native derived word by suppressing the primary stress and using the suffix as the single carrier of prominence, as in: … and make a thyng : … at his writy`ng (GP 325– 326), … in hir dro´nkene`sse : … that I took witne`sse (WBT 381–382). The convention is linguistically motivated only to the extent that derivational suffixes, but not grammatical suffixes, are subject to such metrical promotion. The fashion for iambic versification in Middle English was a cultural import from the Continent, but it could not have been adopted with such ease if the prosodic conditions had not been favorable. The gradual loss of final and inflections in Middle English resulted in a growing number of words realized as monosyllables, allowing flexibility in the prosody-to-meter matching. Increased use of prepositions compensating for inflectional loss created new W S clitic groups: at nı´ght, to re´st, with che´er. Prefixed verbs and adverbs supplied another set of natural iambic structures: befo´re, forgı´ve, perfo´rm, asle´ep. Phrasal stress continued to be right-strong; phrases made up of stressed monosyllables easily match an iambic foot: five bo´oks, tall me´n, full gla´d, God kno´ws. The poets also draw from an inventory of handy “fillers”, semantically dispensable monosyllabic words, e.g. and, now, for, some, and the grammatically redundant “pleonastic” this, that. Thus, although individual underived words retained root-initial stress, in connected speech metrical W S cadences were frequent and easy to construct; this permits an effortless “fit” between language and meter.
8 Linguistic Levels: Prosody Except for the metrical conventions at the line edges, iambic verse provides a reliable framework for reconstructing the stress of Middle English words on the basis of meter-to-prosody correspondences.
5.2 Native and non-native word stress in Middle English The continuing stability of the GSR, aligning primary stress with the left edge of all words and with the left edge of the root for prefixed verbs and adverbs, is easily demonstrated in verse, as in dro´ppyng, ho´uses, smo´ke, chı´dyng, wy´ves, ma´ken in Chaucer’s Wife of Bath’s Tale (WBT ), 278–279: Thow seyst that dro´ppyng ho´uses, and eek smo´ke, And chı´dyng wy´ves ma´ken men to flee
Words derived by suffixation also show the expected main stress on the leftmost root syllable, as kı´ngship, wı´sdom, wı´tness, ho´ly, blı´ssful: The ho´oly blı´sful … (GP 17). Again predictably, the first syllable of compounds is regularly aligned with a metrical S: … for a´ny le´checra`ft (KnT 2745), with wı´lde tho´nder-dy`nt … (WBT 276), to be´ me wa´rde-co`rs … (WBT 359). Compounds usually start in even (S) positions, but since both roots carry a degree of prominence, if the first part is monosyllabic, it can be ´ f clooth-ma´kyng she hadde placed in W, while the second root is in S, e.g. O ´ ´ swich an haunt (GP 447, headless), He was short-sholdred, brood, a thikke knarre (GP 549). Phrasal stress is not testable in iambic verse if there is a buffer weak syllable between the stressed syllables: of so´ndry fo´lk (GP 25), and ma´de fo´rward (GP 33), ty´me and spa´ce (GP 35). Monosyllabic adjectives in noun phrases do provide some corroboration for continuing right-prominent phrasal stress: ne po´lax, ne´ short kny´f (KnT 2544), Gret swe´ryng is (PardT 631), but the stress-alternating nature of the verse and the availability of optional -e and metrical slot-fillers obscure the picture. As argued in Minkova and Stockwell (1997), there is no good reason to posit any dramatic changes in the prosody of phrasal stress from Old English to Present-day English. Even if we assume a more level phrasal stress in Old English than in Present-day English, the right-hand prominence of Old French and Anglo-Norman would have contributed to the present contour. The introduction of a large non-native component into the vocabulary of Middle English is a central theme in any account of the history of English word stress. The non-native vocabulary of Old English never exceeded about 3%, while the portion of the Romance vocabulary at the end of the Middle English period is estimated at about 25%. Once again, attestations in verse provide our best test for the realization of loanwords in the spoken language. Thus we can safely posit initial stress on seson in: Bifil that in that se´son on a day (GP 19), And eek the lusty se´son of that May (KnT 2484). The word was first attested in English 1340–1370 (OED, Simpson [ed.] 2000–), roughly during Chaucer’s lifetime (1343–1400), yet out of the 15 times Chaucer uses the word in The Canterbury Tales and in Troilus and Criseyde, there is one single attestation of the word in rhyme (… thy declinacion : … tyme and his seson (FrT 1033– 1034)) where one could possibly posit right prominence. Such evidence suggests that a
123
124
II Linguistic Levels metrical promotion to seso`n is not different from the treatment of native writy`ng, witne`sse discussed in Section 5.1, i.e., there is no reason to differentiate between native and borrowed words at the line end. The 13th-century loanword country is used 45 times in CT and in Tr, 21 of which are in rhyme position and are realized as end stressed. Of the 24 line-internal attestations, however, there is not a single example of end stress on the word; they are all of the type illustrated by SumT 1710: A mersshy co´ntree called holdernesse. Such findings lead to a serious methodological amendment to the way of collecting verse data for prosodic reconstruction. As argued in Minkova (2000, 2006), the blanket assumption that the verse-final foot provides reliable information on stress is flawed. When we take rhyme position out of the picture, the rate of assimilation of the foreign prosodic contours to the native stem-initial prominence is significantly faster than has been previously acknowledged. The new Romance words coming into the language after the Conquest could be direct loans from the Classical languages, or they could be coming via Anglo-Norman or Old French. Latin (and Greek-via-Latin) disyllabic words would be stressed initially by default: a´xle, e´rgo, hy´mnal, he´rpes, mo´rtar, stu´por, o´nyx were all borrowed in Middle English. According to the Latin Stress Rule, in words of more than two syllables stress falls on the penultimate syllable if it is heavy, otherwise, on the antepenultimate syllable. The Latin Stress Rule in polysyllabic words is thus weight-sensitive, but since many early Latin borrowings lost their inflectional markers (-a, -(t)is, -us, -um, etc.), the picture was often obscured, thus ju´ncture < junctu¯ra, hu´man < huma¯nus. Anglo-Norman and Old French words were stressed depending on the weight of the final syllable: if heavy, the final syllable attracts stress: author, chaplain, jargon, merchant. Light final syllables are unstressed: able, chambre, piece. Since final syllables containing schwa are unstressable, in initially polysyllabic words like bataille, folye, justice, servise, visage the stress was on the penultimate, as in Latin. The extent to which weight-sensitivity at the right edge of the new words affected the prosody of Middle English has often been overestimated, mostly because of misinterpretation of the verse evidence; see Section 5.1. Both disyllabic and trisyllabic preRenaissance borrowings show a strong tendency of leftward stress-shifting, in conformity with the GSR, as in juncture, human, chaplain, merchant, battle, folly, novice, service. The leftward stress-shift disregards syllable weight; indeed in many cases the stress shifts leftwards from a heavy to a light syllable, as in chaplain, battle, folly, justice, novice. Table 8.1 shows the stress profiles of borrowed disyllabic simplex nouns and adjectives in alliterative and syllable-counting verse; the search ignores attestations in the final foot of iambic verse: Table 8.1: Romance loans in Middle English verse (from Minkova 2006: 114) Text
Tokens
Initial Stress
Sir Gawain and the Green Knight The Siege of Jerusalem Troilus and Criseyde Henryson’s poetry
283 87 266 151
276/97.5% 84/96.5% 223/84% 137/90.7%
Non-initial Stress 7/2.5% 3/3.4% 43/16% 14/9.3%
It is evident that the initial wave of borrowing did not upset the stem demarcation on the left. Verbs in which the prefixation is transparent behave like the native prefixed
8 Linguistic Levels: Prosody verbs discussed in Section 4.2.: Perfo´urme it out … (Tr III 417), ye na´t disco´vere me (MerT, 1942). Prefixed nouns and adjectives vary. Chaucer uses both initial and final stress on proverb, a word first recorded in his works (OED, Simpson [ed.] 2000–): Wel may that be a pro´verbe … (WBT 284), And therfore this prove´rbe is … (RvT 4319). Etymologically non-transparent prefixed nouns and adjectives tend to follow the native rule: Ben humble su´bgit … (Tr II 828), … in joye and pe´rfit heele (KnT 1271). The history of stress on prefixed loanwords in Middle English is an area which has not been fully researched yet – it is an inquiry that promises to throw light on the continuity and/or reintroduction of functional stress-shifts in English of the type a´bstract (N, ADJ)–abstra´ct (V); re´cord (N)–reco´rd (V). In iambic verse, polysyllabic words may be hard to fit to a metrical frame of alternating prominences. In Section 5.1 we noted how native suffixes appear to acquire metrically induced secondary stress: bo´ldely`, dro´nkene`sse, do´utele`es, ma´rtyrdo`m. The combination of dominant word-initial stress and the rhythmic preference for stress alternation in borrowed words produces a comparable effect in the new Romance vocabulary: the linguistic /w w s/ in the source language is realized in English as /s w s/: a`moro´use, cha`rite´e, la`xatı´f, o`pposı´t, o`riso´un, ple`ntevo´us, re`gio´un, only in this case it is probably the left edge of the word that carried secondary stress at first, judging from the strong preference for placement of such words in rhyme position: wro´oth was she´ : cha`rite´e (GP 451–452), whı´t : o`pposı´t (KnT 1893–1894), ho´us : ple`ntevo´us (GP 343–344), ado´un : re`gio´un (KnT 2081–2082). The switch from word-initial secondary to primary stress on such trisyllabic words probably started during Middle English, but the precise dating is not recoverable from iambic verse, where both primary and secondary stresses may fill S-positions. The preservation of some degree of stress on the final syllable in Romance loans comfortably beyond Middle English is well documented in Early Modern English, see Dobson (1957: 830–860). The placement of secondary stress on the initial syllable of four-syllable words with an unstressable final syllable: dı`gestı´ble : Bı´ble (GP 437–438) sa`crifı´ce : wı´se (KnT 2369– 2370), dy´e : of bı`gamy´e (WBT 85–86) is also attributable to the principle of rhythmic stress alternation enhanced by the native left-edge prominence pattern. If the final syllable is stressable, the additional stress appears on the second syllable to the left: relı`gio´un : to´un (GP 477–478), coma`ndeme´nt : yse´nt (KnT 2869–2870). The Middle English stress alternation and the eventual demotion and loss of the original primary stress in the foreign vocabulary was attributed to the school pronunciation of Latin in Middle English by Danielsson (1948: 26–29, 39–54) who used the term countertonic accentuation to describe the shift of e.g. post-Classical Latin melancholı´a (1375) to me´lancho`ly, in line with the native model of ma´idenho`d, dro´nkene`sse.
6 Post-Middle English prosodic innovations The rise of literacy in Early Modern English was accompanied by a parallel rapid expansion of the lexicon. Barber (1997: 219–220) estimates that as many as 95 new words were recorded in English during each decade between 1500 and 1700; his counts are based on sampling entries in the OED. This exceeds by far the rate of borrowing in Middle English, which he estimates at 17 new words per decade, using the same methodology. Two-thirds of the new forms in Early Modern English were based on already recorded roots and affixes and about one third were straight borrowings. The large
125
126
II Linguistic Levels majority of these words were coined or adopted by English speakers who were proficient in Latin and Greek and who would therefore automatically apply the Latin Stress Rule to the novel “English” forms: abla´tion, cathe´dral, demo´cracy, mea´nder, te´rminus. The density of these forms and the shared literate understanding of their prosody gave rise to a new, parallel model of stress in English, which is weight sensitive, and ´ regon (1765, possibly Connecticut pidgin which can apply to new words such as O Algonquian, OED, Simpson [ed.] 2000–), kaı´nga ‘village’ (Maori, 1820), palachı´nka ‘pancake’ (Slavic, 1884). The tenacity of the GSR continued during the Early Modern English period in spite of the unprecedented influx of foreign loans, however. Consolidation of the primary stress on the initial syllable of the stem went beyond the disyllabic shifts recorded in Section 5.2 and affected trisyllabic nouns and adjectives: a´morous, cha´rity, la´xative, o´pposite, o´rison, ple´nteous, re´gion have changed their Chaucerian pronunciation in accord with the GSR, similarly ´ınfantry, me´rcury, o´rient, ca´lendar, ge´nial. Stress shift to the initial syllable often proceeds in spite of the etymological heaviness of the penultimate syllable, as in the early loans a´morous, fo´rtunate, ´ınfantry, ´ınterval, o´rient and many post-Middle English forms such as ve´rtebra (1615), ta´lisman (1638), sy´nergy (1660), Ca´vendish (1839), ba´dminton (1845), a´llergy (1911), bo´ondo`ggle (1935). The emerging picture is complexly layered: the prosody of native words follows the Old English left-alignment of word or stem with the main stress. The non-native vocabulary displays hybrid patterns, and no single model covers all realizations without exceptions, so we can only define strong tendencies. New words can fall in with the native left-strong Germanic model, or they can follow a weight-sensitive model whereby stress in non-derived words is assigned by word class and by syllable weight. Verbs with heavy final syllables are generally end-stressed, e.g. para´de, deny´, mainta´in, oblı´ge, prote´ct. Nouns may be stressed depending on the weight of the penultimate syllable in accord with the Latin Stress Rule: age´nda, ca´nopy, horı´zon, ´ınfidel, Toro´nto. Although the considerable overlap between the patterns noted in Section 5.2. for disyllabic nouns and adjectives continues, end-stressed nouns like aby´ss, baro´que, caba´l, cana´l, dure´ss, elı´te, mala´ise, ravı´ne do occur. The extent to which such words retain their prosodic “foreignness” may vary in British and American English. Table 8.2 shows some examples with first attestation dates from the OED; some of these are simply “majority” pronunciations in variation with the alternative pronunciation. Table 8.2: Stress differences between American and British English American English
Date
British English
´ınquiry po´lice (also polı´ce) fru´strate premı´er mo´ustache debrı´s cafe´ gara´ge
(1440) (1450) (1447) (1500) (1585) (1708) (1802) (1902)
inquı´ry polı´ce frustra´te pre´mier mousta´che de´bris ca´fe ga´rage
8 Linguistic Levels: Prosody The hybridity of the Present-day English stress system is also evident in the variability of stress patterns within the last century. Bauer (1994: 96–103) records items which have undergone a recent shift to penultimate stress, e.g. a´bdomen, a´cumen, a´nchovy, e´tiquette, moly´bdenum, pre´cedence, qua´ndary, se´cretive, so´norous, va´gary. He notes a further complicating factor: stress placement in derived words can ignore the nature of the suffix and preserve the prosody of a pre-existing and frequently used base, thus ca´pital, prefe´r are the bases which trigger the change of old capı´talist to current ca´pitalist, and of old pre´ferable to prefe´rable. As noted in Section 4.2, suffixation in Old English was never associated with mainstress reduction; the highest level of prominence for derivational suffixes was secondary stress. The adoption of a large number of foreign affixed words along with their prosodic contours changed this situation. Present-day English suffixes can attract main stress themselves or they can trigger the placement of main stress on one or two syllables to the left of the suffix. Among the suffixes attracting primary stress and reducing the original stress of the base to secondary stress are: -ette (1849), as in ma`jore´tte, -een (1551) as in ve`lvete´en, -ese (1898), as in jo`urnale´se, -eer (1704) as in mo`untaine´er. Final main stress appears also on word endings that may not be etymologically productive suffixes: -ade as in le`mona´de, -ique as in boutı´que, -oo as in ka`ngaro´o. Main stress usually falls on the syllable immediately preceding the suffixes -ic, -id, -ion, -ity/-ety: nume´ric, caro´tid, rebe´llion, tranquı´lity. Among the borrowed suffixes that place the main stress on the antepenultimate syllable of the derived word are -acy, -ast, -ose, -tude: demo´cracy, ico´noclast, co´matose, simı´litude. The antepenultimate is stressed also in combining forms such as -o´logy, -o´sophy, -o´graphy, -o´latry, -o´cracy etc. These new patterns of stress-assignment extend to native roots under foreign suffixation as in Icela´ndic (1674), weathero´logy (1823), speedo´meter (1904), Cha`pline´sque (1921). The placement of stress in derived words has been the subject of much linguistic research. A good descriptive coverage is found in Fudge (1984); the analytical problems are addressed in Giegerich (1999). Another innovation in the post-Renaissance period is the growing productivity of functional stress-shifting in homographic pairs, the a´ddict (N)–addı´ct (V), pre´sent (N)– prese´nt (V), po´lice (N)–polı´ce (V) model, where the shift of stress from one part of speech to another is no longer a matter of prefixation, as in the native shifts in u´pset (N), o´verhang (N). The new stress-shifts do not require compositionality; on the other hand, they are directional (right-to-left) and subject to syllabic and segmental restrictions on the base, not applicable to the native pairs (Minkova 2009). In conclusion, the prosody of Present-day English presents a mixture of word-stress patterns, some inherited from Old English, some introduced in Early Modern English. What we share with Old English is an uninterrupted line of left-edge marking of compounds, unstressable function words and head-prominence in clitic groups, and righthand phrasal prominence. Many relevant details in the prosodic history of English remain under-researched: we lack good documentation of the prosodic behavior of borrowings in Middle English and we still need to evaluate the relevance of competing factors such as phonological composition, frequency, morphological marking and transparency, social prestige, spelling. The relationship between innovations in verse form and prosodic innovations is also of considerable linguistic and cultural interest. Other areas that invite further inquiry are the prosodic patterns in the regional and ethnic varieties of English, and the contact-induced changes in the English spoken in countries where it is an official second language.
127
128
II Linguistic Levels
7 References Barber, Charles. 1997. Early Modern English. Edinburgh: Edinburgh University Press. Bauer, Laurie. 1994. Watching English Change. London: Longman. Benson, Larry D. (ed.). 1987. The Riverside Chaucer. 3rd edn. Boston: Houghton Mifflin. Bredehoft, Thomas. 2005. Early English Metre. Toronto: University of Toronto Press. Danielsson, Bror. 1948. Studies on the Accentuation of Polysyllabic Latin, Greek, and Romance Loan-Words in English: with special reference to those ending in -able, -ate, -ator, -ible, -ic, -ical, and -ize. Stockholm: Almqvist and Wiksell. Dobson, Eric J. 1957. English Pronunciation 1500–1700. Vol. II. Oxford: Clarendon Press. Fudge, Eric. 1984. English Word-Stress. London: George Allen & Unwin. Fulk, Robert D. 1992. A History of Old English Meter. Philadelphia: University of Pennsylvania Press. Getty, Michael. 2002. The Metre of Beowulf. A Constraint-Based Approach. Berlin/New York: Mouton de Gruyter. Giegerich, Heinz. 1999. Lexical Strata in English. Morphological Causes, Phonological Effects. Cambridge: Cambridge University Press. Giegerich, Heinz. 2009. Compounding and Lexicalism. In: Rochelle Lieber and Pavol Sˇtekauer (eds.), Handbook of Compounding, 178–200. Oxford: Oxford University Press. Hutcheson, Bellenden Rand. 1995. Old English Poetic Metre. Cambridge: D. S. Brewer. Minkova, Donka. 2000. Middle English prosodic innovations and their testability in verse. In: Irma Taavitsainen, Terttu Nevalainen, Pa¨ivi Pahta, and Matti Rissanen (eds.), Placing Middle English in Context, 431–461. Berlin/New York: Mouton de Gruyter. Minkova, Donka. 2006. Old and Middle English Prosody. In: Ans van Kemenade and Bettelou Los (eds.), The Handbook of the History of English, 95–125. Oxford: Blackwell. Minkova, Donka. 2008a. Review of Thomas Bredehoft, Early English Metre. Speculum 83(3): 673–675. Minkova, Donka. 2008b. Prefixation and stress in Old English. Word Structure 1(1): 21–52. Minkova, Donka. 2009. Continuity or re-invention in functional stress-shifting. Paper presented at ICEHL 15, Munich. Minkova, Donka and Robert Stockwell. 1997. Against the emergence of the nuclear stress rule in Middle English. In: Jacek Fisiak (ed.), Studies in Middle English, 301–335. Berlin/New York: Mouton de Gruyter. Minkova, Donka and Robert Stockwell. 2005. Clash avoidance in morphologically derived words in Middle English. (Why [-hʊd] but [-dəm])? In: Nikolaus Ritt and Herbert Schendl (eds.), Rethinking Middle English. Linguistic and Literary Approaches, 263–280. Bern: Peter Lang. Russom, Geoffrey. 1987. Old English Meter and Linguistic Theory. Cambridge: Cambridge University Press. Selkirk, Elisabeth O. 1984. Phonology and Syntax: The Relation Between Sound and Structure. Cambridge, MA: MIT Press. Sievers, Eduard. 1893. Altgermanische Metrik. Halle: Niemeyer. Simpson, John (ed.). 2000–. The Oxford English Dictionary. 3rd edn. online. Oxford University Press. www.oed.com/ Stockwell, Robert and Donka Minkova. 1997. Prosody. In: Robert Bjork and John Niles (eds.), A Beowulf Handbook, 55–85. Lincoln: University of Nebraska Press.
Donka Minkova, Los Angeles (USA)
9 Linguistic Levels: Morphology
129
9 Linguistic Levels: Morphology 1 2 3 4 5
Introduction Morphological typology Modern English History References
Abstract Modern English morphology is the result of a long-range typological restructuring, triggered by phonological changes in connection with the emergence of the Germanic language family, leading to an erosion of unstressed final syllables. As a result, the originally root-based morphology became stem-based and finally word-based. Also morphology was originally characterized by pervasive phonologically conditioned morphophonemic alternations, which gradually became morphologically conditioned, because of phonological changes. This was replaced by a simplified system with base invariancy and phonologically conditioned alternations of inflectional endings as a default case characterizing the regular inflection of nouns, verbs and adjectives. The irregular patterns continue properties of the original system and can be interpreted as stem-based with morphologically conditioned alternations of the base form. This is also true of many non-native word-formation patterns, which have been borrowed from stem-based languages such as French, Latin or Greek and have re-introduced base alternation into English derivational morphology.
1 Introduction Modern English morphology is the result of several millennia of linguistic change, which has transformed its ancestor, Indo-European morphology, into a completely different morphological type. These changes often left relics behind, which like tombstones commemorate earlier stages, as e.g. irregular verbs of the type write : wrote : written, which go back to Indo-European ablaut alternations, or irregular noun plurals like mouse : mice, goose : geese, oxen, which reflect old Indo-Germanic and Germanic inflectional classes. Similarly, the stress alternation between verbal o`verflo´w and nominal o´verflo`w, etc. can be traced back to a combination of the Germanic innovation of initial stress and syntactically governed stress distribution in Proto-Germanic (cf. Minkova 2008a, 2008b), eventually acting as a landing site for Romance loans like reco´rd V and re´cord N. This new Germanic stress system is also held responsible for the loss of unstressed syllables, many of which were carriers of morphological information, the result being a language with little inflection left. On the other hand, there are also typological innovations like the vowel and/or consonant alternations in sane : sanity, serene : serenity, Japa´n : Ja`pane´se, hı´story : histo´ric : hı`storı´city, ele´ctric : ele`ctrı´city, close : closure, resulting from the integration of non-native (Romance, Latin and Neo-Latin) wordformation patterns into English with a concomitant variable stress system, which reverses the original typological drift towards a non-alternating relation between bases and derivatives and is reminiscent of Indo-European, where variable stress/accent Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 129–147
130
II Linguistic Levels produced variable vowel quality/quantity (ablaut). It is such long-range developments and their typological consequences that will be dealt with in the following. The existing literature, e.g. Brunner (1960–62), Wełna (1996), the respective chapters in Hogg (1992), Blake (1992), or Hogg and Denison (2006), pay only cursory attention to these typological changes. Moreover, they reflect the Neo-grammarian tradition of looking backwards to the Indo-European and Germanic morphological structures, interpreting later patterns in terms of older structures as having undergone losses rather than as acquiring structural innovations. As a consequence, morphological restructuring is often not topicalized enough or assumed to have occurred much later than it actually did. A good example is Brunner’s (1962: 3–10) treatment of Old English nominal inflection. He makes a point of describing this in terms of Indo-European inflectional classes involving stem-formatives like IE -o- (= Grmc. -a-), IE -o¯- (= Grmc. -a¯-), etc., arguing that reference to the Indo-European categories rather than the Germanic ones is preferable in order to emphasize the fact that the Germanic declinations correspond exactly to the declinations of other Indo-European languages. The synchronic relevance of classifying Old English nouns in terms of Indo-European o- or a¯-stems, etc. is simply taken for granted and never questioned, which obscures a profound typological change between Indo-European and Old English, viz. the shift from root-based to stem-based and even word-based morphology and the demise of the stem-formatives as a functional category. Even though this approach allows for restructurings, it would tend to place them at a later period than when they actually occurred. What is needed, therefore, is a Janus-like approach: this should take into account both where a synchronic linguistic stage comes from (= retrospective), but also in which direction it is moving (= prospective).
2 Morphological typology 2.1 Traditional typology Morphological typology as introduced by the Schlegels (Schlegel, Friedrich 1808; Schlegel, August W. 1818) and Humboldt (1827–29) covered only inflectional morphology and was based on two overlapping scales with one parameter each. The first parameter resulted in the distinction between analytic languages, where grammatical functions are expressed word-externally by prepositions, auxiliaries, etc., and synthetic languages, where grammatical functions are expressed word-internally by inflection, incorporation, etc. The second parameter was based on how grammatical functions are represented formally and resulted in the distinction between isolating, agglutinating, inflectional, and incorporating languages. In isolating languages, grammatical functions are expressed word-externally; i.e. this type coincides with the analytic one, and languages belonging to this type have no inflectional morphology. The other three are sub-types of the synthetic type. In agglutinating languages, there is a one-to-one correspondence between a grammatical category and its exponent, as, e.g., in Turkish, Finnish or in English, where number and case are expressed consecutively, if the plural is irregular, cf. (1): (1)
ox en s
geese s
PL GEN PL
GEN
boy s
Ø
PL GEN
9 Linguistic Levels: Morphology In inflectional languages, one morphological exponent represents more than one grammatical category, functioning as a portmanteau morph, cf. Lt. am-o¯ ‘I love’ (-o¯ = 1P [person] SG [number] PRES [tense] IND [mood] ACT [voice]), or OE cyning-as (-as = PL [number] + NOM/ACC [case]). In incorporating languages, grammatical functions such as subject, object, adverbial complement are integrated into the predicate itself together with their nominal carriers; this might be postulated for recent English formations such as machine-translate, thought-read, flight-test, etc. (cf. Section 3.2.2.)
2.2 Extended typology These parameters only partially characterize the overall gestalt of the morphology of a language and therefore have to be supplemented by further aspects, which will lead to a more differentiated picture. The following additional parameters have proved useful (cf. Kastovsky 1997, 2006a, 2006c): – – – – – –
morphological status of the input to the morphological processes (word, stem, root) number and status of inflectional classes formal representation of inflectional and derivational markers status and function of morphophonemic/allomorphic alternations position of affixes/position of the head existence and status of morphological levels (e.g. native vs. foreign)
The first parameter requires a comment. I will adopt the distinction between “lexeme” (= dictionary entry), “word-form” (inflectional form of a lexeme), and “word” (actual representation of a lexeme via word-form in an utterance as independent syntactic element or free form) from Matthews (1974: 20–26) and Lyons (1977: 18–25). “Word-formation” (“derivational morphology”) is therefore actually “lexeme-formation”, but I will retain the former term for convenience’s sake. What is crucial is the status of the lexeme representation which acts as input to inflectional and derivational processes, the “base form”. This is defined as that lexeme representation from which all word-forms and the result of word-formational processes can be derived. Thus, word-formation and inflectional morphology are interdependent: the input to word-formation processes, and the demarcation between inflection and derivation, depends on the typological status of inflection. For this, the following distinction is suggested: a) word-based morphology: The base form can function as a word (free form) in an utterance without the addition of additional morphological (inflectional or derivational) material, e.g. ModE cat(-s), cheat(-ed), beat(-ing), sleep(-er). b) stem-based morphology: The base form does not occur as an independent word, but requires additional inflectional and/or derivational morphological material in order to function as a word. It is a bound form (= stem), cf. OE luf-(-ian, -ast, -od-e, etc.), luf-estr-(-e) ‘female lover’, Grmc. *dag-(-az) ‘day, NOM SG’, ModE scient-(-ist) vs. science, dramat-(-ic) vs. drama, astr-o-naut, tele-pathy; thus luf-, luf-estr-, *dag-, dramat-, astr-, -naut, tele-, -pathy are stems. c) root-based morphology: Here the input to morphological processes is even more abstract and requires additional morphological material to become a stem, to which the genuinely inflectional endings can be added in order to produce a word.
131
132
II Linguistic Levels Such roots can either be affiliated to a particular word-class, or they can be word-class neutral. In this case the word-class affiliation is added by a word-formative process, cf. IE roots like *wVr- ‘bend, turn’ (cf. Lt. uer-t-ere ‘turn’, OE weor-þ-an ‘become’, wyr-m ‘worm’, etc.), with V standing for the ablaut vowel, whose shape is determined by the morphological process in question. Whether Indo-European roots were word-class specific or neutral is not quite clear, but not relevant for our purposes. Terminology varies considerably in this respect, cf., e.g., Bauer (1983: 20–21; 1992: 252– 253), Giegerich (1999) or Huddleston and Pullum (2002: 1624–1625). The terminology used here is geared to the history of Indo-European morphology, with its shift from a root-based to a stem-based and a (partly) word-based system. I will therefore use the term “stem” for what some linguists call “root” also with reference to English, i.e. for a lexical element which is bound and which can only occur as a word with additional morphological material, as, e.g., scient-, dramat-, tele-, -pathy. The term “root” will be restricted to the Indo-European period and is not relevant for the Indo-European daughter languages. The ultimate starting point of English was a root-based morphology (IndoEuropean), which became stem-based in the transition to Germanic. In the transition from Germanic to Old English, inflection became partly word-based, and this eventually became the dominant typological trait of Modern English.
3 Modern English 3.1 Inflectional morphology 3.1.1 Inflectional morphology vs. word-formation In Modern English, there is a fairly neat division between inflectional morphology and word-formation, except for adverb formation in -ly, which is ambivalent. Since -ly produces a change of word-class (ADJ > ADV), it is usually treated as derivational. But on account of the complementary distribution of adverbs and adjectives as in (2) (2)
He smokes heavily : He is a heavy smoker.
it could also be treated as inflectional, parallel to nominal case inflection, cf. Hockett (1958: 211). This is probably why Marchand (1969) did not include adverb formation in his handbook.
3.1.2 Regular vs. irregular inflectional morphology Inflectional morphology of verbs and nouns is based on a default system consisting of two inflectional classes: regular and irregular. Regular inflection is fully predictable. It is word-based, base-invariant, and the morphophonemic alternations of the inflectional endings are phonologically conditioned: 3P SG, plural and genitive have the allomorphs: /z/ (underlying form and default), e.g. kids /kɪdz/; /ɪz/ (vowel insertion after base-final coronal sibilants), e.g. kisses /kɪsɪz/; and /s/ (devoicing after base-final voiceless consonants), e.g. bets /bets/. Preterit and past participle have the allomorphs: /d/ (underlying and default), e.g. loved /lʌvd/; /ɪd/ (vowel insertion after base-final alveolar stops), e.g. hated /heɪtɪd/; and devoicing after base-final voiceless consonants, e.g. kissed /kɪst/.
9 Linguistic Levels: Morphology Irregular verbs and nouns can be grouped into various sub-classes (classification in the grammars varies considerably); they have either variable bases or phonologically unpredictable inflectional allomorphs, or both, and are relics of historically earlier stages, cf. (3a, b): (3)
a. keep /ki:p/ : kep-t /kep-t/ (variable base, regular inflectional allomorph), deal /di:l/: deal-t /del-t/ (variable base and unpredictable inflectional allomorph /t/) shut : shut-Ø (invariable base, unpredictable inflectional allomorph Ø), write : wrote : writt-en, sing : sang : sung, bleed : bled : bled (unpredictable inflectional allomorphs, viz. replacives, -en); b. leaf /li:f/ : leav-es /li:v-z/ (variable base, regular inflectional allomorph), ox : ox-en (invariable base, unpredictable inflectional allomorph), mouse : mice, child : child-ren, fung-us : fung-i (unpredictable inflectional allomorphs)
The status of /kep/, /del/, /li:v/ /ʧɪld/ is arguable. On the one hand, these are allomorphs of the lexeme representations /ki:p/, /di:l/, /li:f/, /ʧaɪld/, which are words. But since they only occur together with inflectional morphemes, these allomorphs might be regarded as stems, i.e. these lexemes have both a word-based and a stem-based morphological representation. This is similar to the derivational pattern science : scient-ist, drama : dramat-ic, where the lexical base is represented as a word, but also has an allomorph which occurs in derivatives and which has to be regarded as a stem. Write : wrote : written, or mouse : mice are word-based with a replacive morph representing the inflectional morpheme, which in the case of writt-en might be interpreted as discontinuous, consisting of the replacive /aɪ →ɪ/ and the suffix -en. In this case, the morphophonemic alternation is morphologically relevant and expresses a morphological contrast, whereas in the cases of keep : kept, leaf : leaves, etc. the morphophonemic alternation is non-functional. Synchronic classifications do not always reflect historical developments. Thus, Modern English sing : sang, write : wrote and bleed : bled all have replacives as tense and participle markers, but with sing and write the replacive reflects an Indo-European ablaut alternation, whereas with bleed : bled the alternation has the same origin as that occurring in keep : kept, deal : dealt, viz. shortening of long vowels before certain consonant clusters in late Old English. But with bleed : bled the alternation has become morphologically distinctive because of the lack of any other overt exponent of tense/participle, whereas the alternation in keep : kept might be treated as morphologically non-distinctive. On the other hand, bleed : bled and shut : shut share the same history with regard to the loss of the original past tense/past participle ending, cf. OE ble¯d-d-e, ge-ble¯d-(ed), scyt-t-e, ge-scyt-t-(ed): the ending was lost (= replaced by zero, cf. Kastovsky 1980), when gemination was lost in Middle English; but with bleed : bled, a formal contrast had arisen thanks to vowel shortening in the preterit/ past participle, which was morphologized, whereas with shut no formal contrast remained, which results in a zero allomorph for preterit and past participle.
3.1.3 Inflectional morphology of nouns For nouns, the following morphological categories are relevant: number, case, class. Gender does not figure as an inflectional category but is determined by the nature of the referent (= natural gender), in contradistinction to Old English, where it was a
133
134
II Linguistic Levels relevant grammatical category in the organisation of inflectional paradigms. For its loss, cf. Kastovsky (1999). The category of number is primary; the category of case (genitive) is secondary and dissociated from number, i.e. has a separate exponent, in contradistinction to Old English, where the two categories were fused, cf. cyning-as NOM/ACC PL, cyning-um DAT PL. In view of structures such as the Queen of England’s castles, where the genitive attaches to the whole NP rather than to its head Queen, the genitive might be regarded as a clitic, and not as an inflection. The category of class relates to the formal expression of plural, which involves a default system: regular plurals with base invariance and phonologically determined plural allomorphs, and irregular plurals, which deviate from this, cf. (3b).
3.1.4 Inflectional morphology of verbs For verbs, the following inflectional categories are relevant: tense, number, person, mood, finiteness, class. On the basis of the degree to which these categories are exploited, the verb system can be subclassified into three subcategories, each with distinct properties: 1) modal verbs, 2) lexical verbs, 3) the copula be. Modal verbs are the continuation of the Germanic preterit presents, which have undergone a shift towards a separate category in Middle and Early Modern English. They only exploit the tense dimension, i.e. they have no infinitive or participle forms, no person/number marking and no mood contrast. The verbs in question are can, may, shall, will, must, ought to and, to a certain extent, dare, need. Semantically speaking, the tense contrast is problematic, since the morphological opposition can : could, may : might, etc. most of the time does not really signal ‘past’ vs. ‘non-past’, but rather a different degree of modality. In view of their specific morphological and semantic properties, the modal verbs are treated as a separate category. Lexical verbs are characterized by the contrast between an unmarked base form (which also functions as infinitive and non-preterit), the 3P SG, and the marked preterit, present participle, and past participle. The copula be has the richest morphology with a person/number contrast in the preterit, and person distinctions marking first, second and third person in the non-past. With the lexical verbs, person and number are expressed jointly, but they only have a person/number marker in the 3P SG non-preterit, which might better be regarded as an agreement morpheme rather than a person/number morpheme proper. The morphological dimension of mood involves the categories indicative, subjunctive, and imperative, where the morphological marking is reduced to the absence of the agreement morpheme, whereas in Old English there were full-fledged separate paradigms. Inflectional class is an inherent property of the verb, determines how it forms the preterit and the second participle, and is based on a default system. Regular verbs are the default case: they are base-invariant and select the phonologically conditioned allomorphs /d/, /t/ and /ɪd/. The irregular verbs (c.200) involve a modification of the base form (allomorphy), or select a non-phonologically conditioned representation of the past tense and past participle morphemes or both.
9 Linguistic Levels: Morphology
3.2 Word-formation Modern English word-formation includes the following processes: compounding, affixation (prefixation, suffixation and zero-derivation/conversion), clipping, blending, acronyms, sound symbolism; for a comprehensive, somewhat dated but still relevant survey cf. Marchand (1969). It can be argued (Kastovsky 2009b) that most of these word-formation processes are prototypical patterns arranged on a scale of progressively less independent constituents, cf. (4): (4)
word compounding > stem compounding > clipping and clipping compounds > affixation (word-/stem-based) > blending (splinters)) > acronyms
3.2.1 Compounding English compounds are conventionally defined as combinations of full words resulting in a new lexical unit, e.g. bird cage, flatfish, earthquake, house-keeping, law-breaker, sunrise, writing table, dance hall; color-blind, icy-cold, heart-breaking, easy-going, manmade, etc. These illustrate nominal and adjectival compounds, most of which have been in the language since pre-Old English. The status of verbal compounds is problematic. Marchand (1969: 96) accepts as compounds only verbs with a locative particle like overdo, underestimate, and outdo, but with some reservations because of semantic problems with the first constituents, which make them look like semi-prefixes. These continue Old English formations with inseparable particles originally having a purely locative meaning, which was later extended metaphorically to ‘excess’, ‘deficiency’. For the type stagemanage, playact, spoonfeed, newcreate, Marchand (1969: 104) assumes backderivation from nouns such as stage-manager, playacting, spoonfed, newcreated, since the Germanic languages never had a genuinely productive verbal compound pattern. But in the last decades, the situation has changed, and a verbal compound pattern based on incorporation is apparently taking over, which no longer needs a nominal base as a source, cf. quick-march, new-form, slow-step, slow-kill, dark-adapt, quick-frost, surekill, hard-learn, quick-check; quantum-teleport, knee-jerk, machine-translate, thoughtread, flight-test, depth-bomb, finger-tap, hull-walk, flash-vaporize, flash-freeze, randomly culled from science fiction novels published in the last 15 years. Even V + V compounds are catching on, cf. think-hiss, whisper-hiss, touch-share, wobble-hop, glide-walk, skimglide, bend-swivel (in the form bent-swivelled), strip-search, hop-step, strip-mine from the same kind of source (cf. Wald and Besserman 2002). The compound principle as such is old and goes back to Indo-European, probably even to its oldest layer before the development of certain inflectional and syntactic categories like full relative clauses (cf. Kastovsky 2009a). Therefore, at this stage compounds do not seem to have consisted of full-fledged words, but of roots and stems (with stemformatives). At a later stage syntactic groups consisting of full-fledged words with case marking of the first member expressing its semantic-syntactic relation to the second member seem to have coalesced to compound patterns, carrying along their original first-member inflectional endings. These eventually lost their syntactic function and became mere morphological linking elements like the original stem-formatives, cf. ModE driver’s seat, bull’s eye or Ger. Universita¨tsbibliothek, Kindergarten, Frauenkirche.
135
136
II Linguistic Levels
3.2.2 Stem-compounds Formations such as astr-o-naut, Mars-naut, astr-o-physics, agr-o-chemical, agr-i-culture; hyper-active, hyper-aemia, omn-i-pre´sent, omn-ı´-scient, hepat-itis, wakeup-itis, astr-ology, ex-o-bio-logy, megal-o-mania, star-mania belong to the non-native level of English word-formation. Their constituents have been dubbed “combining forms” in the Oxford English Dictionary for want of a better term at the time of its inception, but there has never been a satisfactory definition of this heterogeneous category (cf., e.g., Prc´ic´ 2005, 2007, Kastovsky 2009b). This term is misleading, covering a number of different phenomena, and should therefore be discarded. Instead, we should recognize the existence of stem-compounds in English, i.e. compounds whose constituents are lexeme representations without word-status, a pattern inherited from languages like Latin or Greek. Formations such as astr-o-naut, agr-i-culture, omn-i-scient, etc. thus contain stems, and the middle element is – historically speaking – a stem-formative, but today functions as a linking element as is the case with -s- in word-compounds. Elements like -logy also go back to such stems, but have developed into suffixes, just as OE do¯m in kingdom, stardom or OE ha¯d in childhood, statehood, where do¯m, ha¯d had originally been independent words which developed into suffixes. Similarly, elements like ante-, anti-, auto-, mono-, etc., originally independent lexemes in the source languages, have developed into prefixes. And the elements Ameri- or Euro- in Euro-City can be regarded as clipped stems (< Ameri(ca), Euro(pa)), which can also be used in compounding.
3.2.3 Affixation Let me now turn to affixation. There are three basic sources for affixes: a) bleaching and grammaticalization of lexemes as compound constituents, as in -dom, -hood, or for-, be- and many other prefixes going back to locative particles; b) borrowing from other languages, as was the case with the majority of ModE prefixes (e.g. ante-, auto-, co-, counter-, de-, dis-, ex-, hyper-, in-, mono-, non-, post-, pre-, re-, sub-, and many others) and also suffixes (e.g. -able, -age, -al, -ance, -ation, -ery, -ess, -ify, -ism, -ist, -ity, -ize, -ment, -ory, and many others); c) secretion, as in -teria caketeria, washeteria), -gate < Watergate (> Irangate, zippergate), -burger < hamburger (> beef-burger > burger), etc. The latter two sources, borrowing and secretion, are closely related, because the borrowing of a non-native affix also involves secretion, since the borrowed affixal element has to be recognized as a separate morphological unit before it can become productive. The demarcation between compounding and affixation is fluid because of the diachronic shift from lexeme to affix, resulting in a bridge category “semi-affix”, e.g. -monger, -like, -wise (Marchand 1969: 356–358), or out-, over-, under- in their metaphorical meaning. On the other hand, affixes might also turn into lexemes, cf. burger, mini, extra.
3.2.4 Zero-derivation There is one word-formation process whose status is controversial, but which I interpret as a sub-type of suffixation, i.e. derivation by means of a zero morpheme, as in salt N >
9 Linguistic Levels: Morphology salt V, clean ADJ > clean V, cheat V > cheat N, walk V > walk. Several approaches have been suggested to account for this process; the following two seem to be the most prominent ones. The older is “conversion”, which assumes that this process is different from normal suffixation and simply consists of converting a lexeme belonging to one word class, e.g. salt N, to a lexeme belonging to another word class, salt V. Connected with this is the view that this is a recent phenomenon in English, which only became really productive after the loss of inflectional endings (cf., e.g., Biese 1941). The second approach assumes that this is a regular derivational process following the binary determinant/determinatum schema of word-formation, except that the derivational element is not overtly present, but is there in content; this is formally expressed by assuming a zero morpheme as the derivational element, which allows one to account for the semantic addition involved in this process. It is parallel to suffixal derivation, cf. carbon N > carbon-ize V ‘provide with carbon’ = salt N > salt-Ø V ‘provide with salt’, black ADJ > black-en V ‘make black’ = clean ADJ > clean-Ø V ‘make clean’, teach V > teach-er N ‘someone who teaches’ = cheat V > cheat-Ø N ‘someone who cheats’, walk V > walk-ing N ‘action of walking’ = walk V > walk-Ø N ‘act of walking’ (Marchand 1969: 359–364). This interpretation presupposes a neat distinction between inflection and derivation, where inflectional endings have no derivational force, and therefore allows us to treat as zero-derivations also instances where the derived word is accompanied by inflectional affixes, as in OE a¯rN ‘honor’ > a¯rN-ØV-ianINF ‘to honor’, de¯opADJ ‘deep’ > de¯opADJ-ØV-ianINF ‘to deepen’, feohtV-anINF ‘to fight’ > feohtV-ØN-eNOM.SG‘fight’. The latter approach, apart from taking into consideration the semantics of the derivational process, also has the advantage of reflecting the historical development, because in all the instances where today we would have to assume zero-derivation, there originally was some derivational marker in the guise of a stem-formative, which was lost due to phonetic attrition and was “replaced by zero” (cf. Kastovsky 1980, 1996).
3.2.5 Blending The end of the scale is made up by blending, which is a heterogeneous category and consists of formations where both parts consist of curtailed lexemes, e.g (5a, b, c). (5)
a. Chunnel = Ch-(annel) + (t)-unnel motel = mo-(tor[ist]) + (ho)-tel b. Oxbridge = Ox-(ford) + (Cam)-bridge transceiver = trans-(mitter) + (re)-ceiver c. smog = sm-(oke) + (f)-og brunch = br-(eakfast) + (l)-unch slithy = sli-(m) + (li)the + y chortle (chuckle + snort)
Examples (5a, b) echo the structure of regular determinative compounds, except that both constituents are clipped. Oxbridge, transceiver are clipped dvandva compounds, which denote a combination (union) of the respective referents, i.e. Oxbridge ‘the universities of Oxford and Cambridge’ is comparable to Austro-Hungary ‘an entity which consists of both Austria and Hungary’, similarly concavo-convex. This compound type is an innovation in English, borrowed from Latin and Greek. The formations in (5a, b) are
137
138
II Linguistic Levels usually treated on the same footing as (5c), but this does not seem to be correct. While the examples in (5a, b) echo regular compound structures, those in (5c) do not. They denote a referent which is a mixture or crossbreed of the referents of the constituent lexemes. They have no direct counterpart in compounding and seem to be a relatively recent development alongside the tremendous rise of acronyms like aids, laser, NATO, FBI, CIA, which form the endpoint of the scale in (4).
3.2.6 Typological properties Let me conclude this section with a few general remarks on the typological properties of Modern English word-formation. The input to Modern English word-formation can be words (at the native level), stems (at the non-native level), both of which can be curtailed (clipped) in clipping-compounds and blends. This implies that there are different morphological levels with different properties. The native level is characterized by base invariancy (no alternations), affix-invariancy (except for phonologically determined alternations) and stress always located on the first constituent (= German stress rule) except for verbal prefixes as in u`ntı´e. The non-native level allows for both stem and affix variation as well as variable stress position coupled with vowel alternation, cf. intolerable : impossible : illegible, Japa´n : Ja`pane´se, etc. As in all Germanic languages, the standard constituent sequence in word-formations is determinant/determinatum (modifier/head). There is one controversial area, however, viz. prefix formations of the type defrost, disarm, unbutton. Some linguists (e.g. Lieber 1983: 253) assume that the prefixes here act as heads in order to account for the change of word-class involved, thus reversing the standard sequence of modifier/ head. This analysis coincides, interestingly, with the rejection of zero derivation as a productive derivational process; rather the relationship between the involved lexical items is described as a static lexical correspondence. But as has been argued above, zero-derivation might be regarded as a sub-type of suffixal derivation, and in the case of the prefixal verbs defrost, disarm, unbutton, the second part might also be regarded as zero-derived, functioning as a head, which would make the prefix a modifier, as befits.
4 History 4.1 Indo-European Some features of Modern English morphology go back to Indo-European, e.g. ablaut like sing : sang : sung, but also the division into strong and weak verbs and the various inflectional classes of nouns. Therefore, a look at Indo-European might be useful. Its central morphological category was the root, a monosyllabic consonantal skeleton with a vowel slot (conventionally assumed to have been /e/ with its ablaut alternants) and certain restrictions on the co-occurrence of consonants in onset and coda position (cf. Clackson 2007: 69–71). The actual nominal, adjectival, or verbal paradigms were derived by adding stem-forming elements and/or other derivational elements again followed by stem-formatives, to which the inflectional endings proper were added, as in (6): (6)
root + stem-formative (+ derivational affix + secondary stem-formative elements) + inflection proper
9 Linguistic Levels: Morphology From such roots primary nouns and primary verbs were formed. There was no directional connection between verb and derived noun, or noun and derived verb; both were indirectly related via a common root, from which they were derived by independent morphological processes. This relationship characterizes the ancestors of the strong verbs and the related nouns and adjectives. From these primary derivatives further, secondary derivatives could be formed (see Figure 9.1). Root Primary stem-formation Noun
Verb
Secondary stem-formation Verb
Noun
Figure 9.1: Derivational relationship between IE roots, noun stems, and verb stems
The Indo-European verb system was characterized by a mixture of aspectual and temporal categories such as present, imperfect, perfect, aorist, etc., and mode of action categories such as iterative, intensive, durative, inchoative, etc. In Modern English or German, these categories are separated as either belonging to inflection (tense, aspect) or derivation (mode of action/aktionsart). In Indo-European, however, no such separation was possible, and these aspect/mode of action categories were derived directly from the root, resulting in stems (see Figure 9.2). Root
Present stem
Perfect stem
Aorist stem
Figure 9.2: Verbal stems in Indo-European
Aspect formation and the person/number inflection of these primary verbs were characterized by morphologically-governed stress alternations, which produced nonfunctional differences of vowel quality and quantity, called ablaut. This is the source of the Germanic system of strong verbs, consisting of six classes (and one additional class of reduplicative verbs fitting into the same pattern), which are based on the syllable-structure of the root, viz. ablaut vowel + syllable coda in (7): (7)
Class 1: -ViTClass 2: -VuTClass 3: -VL/NTClass 4: -VL/NClass 5: -VTClass 6: -VH(V = ablaut vowel, T = obstruent, L = liquid, N = nasal, H = laryngeal)
The secondary derived verbs, i.e. those derived from primary nouns, verbs, or adjectives, but not directly from roots, only occurred in the non-perfective (= present) aspect. They are the ancestors of the Germanic weak verbs containing stem-formatives like -j- and -o¯j- characterizing class 1 and 2 weak verbs.
139
140
II Linguistic Levels
4.2 Germanic to Old English: verbal system One major Germanic innovation was a shift from an aspectual to a tense system. This coincided with the shift to initial accent, and both may have been due to language contact, maybe with Finno-Ugric. Initial stress deprived ablaut of its phonological conditioning, and the shift from aspect to tense required a systematic marking of the new preterit tense. For this, two types of exponents emerged. One is connected to the secondary (weak) verbs, which only had present aspect/tense forms. They developed an affixal “dental preterit”, together with an affix for the past participle. The source of the latter was the Indo-European participial -to-suffix; the source of the former is not clear (cf. Tops 1974). The most popular theory is grammaticalization of a periphrastic construction with do (IE *dhe-), but there are a number of phonological problems with this. The second type was the functionalization of the originally non-functional ablaut alternations to express the new category, i.e. the making use of junk (Lass 1990). But this was somewhat unsystematic, because original perfect forms were mixed with aorist forms, resulting in a pattern with over- and under-differentiation. Thus, in class III (helpan : healp : hulpon : geholpen) the preterit is over-differentiated, because the different ablaut forms are non-functional, since the personal endings would be sufficient to signal the necessary distinctions. But in class I (wrı¯tan : wra¯t : writon : gewriten), there is under-differentiation, because some preterit forms and the past participle have the same vowel. This situation proved unstable and was levelled out in Middle English, along with a loss of many strong verbs or their shift to the regular (= originally weak) class. Moreover, the transparent syllable structure on which this system had been based had lost its transparency in Old English due to numerous sound changes. It is therefore arguably possible to treat the Old English strong verbs already as irregular, i.e. to assume that the switch from a system based on the distinction between weak verbs characterized by stem-formatives and strong verbs characterized by ablaut without a stem-formative to a system based on the dichotomy of regular and irregular verbs had already occurred in Old or even in pre-Old English (cf. Kastovsky 1997). The weak verbs were originally derived from nouns, adjectives or verbs by various stem-formatives, which are still recognizable in Gothic, cf. (8): (8)
Class Class Class Class
1: 2: 3: 4:
-j-/-i-o¯j-/-o¯-e¯(j)-no¯-
(Go. (Go. (Go. (Go.
sat-j-an, OE sett-an) salb-oˆ-n, OE sealf-ian) hab-an, OE habb-an) full-n-an, ?OE beorht-n-ian)
Classes 1 and 2 survive in Old English. Remnants of class 3, e.g. libban, lifde ‘live’, habban, hæfde ‘have’, hyc˙g˙an, hogode ‘think’ are best treated as irregular. Class 4 seems to have survived with -n- having been re-interpreted as a derivational suffix, e.g. beorht-nian < OE beorht ADJ, the antecedent of the -en-suffix in blacken, redden, fasten, etc. At this stage, the stem formative had a dual function: it acted as a derivational morpheme like -ize, -ify, -en, but it also had an inflectional function characterizing a particular inflectional class. Gradually, the weak verbs lost their status as a derived category (cf. Kastovsky 1996). Many of them were reinterpreted as basic, partly because their derivational base was lost, partly due to semantic reinterpretation. The ablaut nouns related to strong verbs
9 Linguistic Levels: Morphology
141
usually denoted agents, actions, results or instruments related to the underlying verbal action. These lexical-semantic categories are typically deverbal. Now these deverbal nouns often served as the basis for secondary verbal derivatives, e.g. faran ‘to go, travel’ > fo¯r ‘going, journey’ > fe¯ran ‘go, come, depart’ (= ‘make a journey’). In this way, many pairs of the type action noun > verb ‘perform action’, agent noun > verb ‘act as agent’ came into existence. For these, however, the direction of derivation was just the opposite of the one relating the basic nouns to their strong verbs and in many instances came to be reinterpreted in the same way as in back-derivations such as peddler > peddle ‘act like a peddler’ → peddle > peddler ‘someone who peddles’ because of write > writer ‘someone who writes’. Such reinterpretations eventually established a general pattern of nominal derivation from weak verbs, so that these lost their exclusively derived character, cf. the following examples of clearly deverbal nouns in (9): (9)
hwistlian ‘to whistle’ > hwistle ‘whistle’ cnyllan ‘to strike, knock, ring a bell’ > cnyll ‘clang, stroke of a bell’ huntian ‘to hunt’ > hunta ‘hunter’
While in Germanic and pre-Old English the stem formatives were clearly recognizable throughout the respective paradigms, this was no longer true of Old English, where they have fused with the person/number endings, cf. trymm-an ‘to strengthen’ < trum ‘strong’, lufian ‘to experience love’ < lufu ‘love’, de¯man ‘to pass judgement’ < do¯m ‘judgement’, nerian ‘to save’; see Table 9.1. Table 9.1: Typical paradigms of Old English class 1 and class 2 weak verbs
Class 1
Stem
+
Person/Number/Inf.
trymm trymm trym trym trymm de¯m de¯m de¯m de¯m de¯m ner ner ner ner ner
+ + + + + + + + + + + + + + +
anINF e1P SG est2P SG eð3P SG aðPL anINF e1P SG st2P SG þ3P SG aðPL ianINF ie1P SG est2P SG (e)þ3P SG iaðPL
Class 2
Stem
+
Person/Number/Inf.
luf luf luf luf luf
+ + + + +
ianINF ie1P SG ast2P SG að3P SG iaðPL
The stems of class 1 may exhibit morphologically conditioned allomorphy caused by West Germanic Consonant Lengthening (e.g. trymm-an : trym-est : trym-ed-e); so did the person/number morphemes in both classes. The choice of the latter was determined by the class membership of the verb, which was an inherent feature of the stem, no longer linked to an overt class-defining stem-formative, i.e. external class characterization had become holistic, i.e. paradigm-dependent.
142
II Linguistic Levels In the past tense and the past participle, the two classes still seemed to be kept apart by the reflexes of the original stem formatives /e/ and /o/, as in (10): (10) a. Class 1: trym+e+dPAST+e1P SG trym+e+d+est2P SG trym+e+d+onPL gePP+trym+e+dPP b. Class 2: luf+o+dPAST+e1P SG luf+o+dPAST+est2P SG luf+o+d+onPL gePP+luf+o+dPP
This analysis we find implicitly or explicitly in the traditional handbooks, and it is the primary basis for the distinction of the two classes. But it is questionable synchronically, especially in view of the increased number of class shifts, beginning with the overlap of the nerian- and the lufian-type, but eventually going far beyond it. The morphological structure postulated for the present without a stem-formative must have also been extended to the preterit already in the course of the early Old English period. This was certainly a gradual process, and there was a transition period of morphological indeterminacy. But by the end of the 9th century, if not even earlier (the earliest class shifts might make this date more precise), the change had been complete, and the phonological relics of the stem-formatives had been reinterpreted as part of the underlying representation of the preterit/past participle morpheme, i.e. (11): (11) trym+e+dPAST+e > trym+edPAST+e luf+o+dPAST+e > luf+odPAST+e This was facilitated by the pre-Old English syncopation of /i/ in class 1 when preceded by a heavy stem or dental stops (i.e. Siever’s Law/High Vowel Deletion), cf. de¯m-d-e, set-t-e, as against ner+e+d+e. With fæstan, vowel deletion resulted in an ungrammatical cluster /fæst+t+e/, which was simplified to /st/ with concomitant loss of the representation of the preterit and past participle morpheme, i.e. fæst+Ø+e. This is the first zero allomorph of the preterit/past participle, i.e. the antecedent of the type cut : cut : cut. More instances arose through the loss of geminates in Middle English, when forms like OE set+t+t+e developed into set+Ø(+e). This development had a very important morphological side effect. The stem formatives had had a dual function, derivational and inflectional. With their loss, inflection became lexically determined (implicational), and the derivational function was no longer expressed overtly: overt derivation became zero-derivation. At the same time, the fuzzy delimitation between inflection and derivation became clear-cut. Incidentally, zero-derivation was the only process for creating denominal verbs, since the -n-suffix only derived deadjectival verbs. It was only in Middle and Early Modern English that new suffixes were introduced through borrowing, viz. -ate, -ify, and -ize. Towards the end of the Old English period, the only class whose morphological behavior was fully predictable was class 2. Verbs of class 1 exhibited morphophonemic/morphological alternations, partly old, partly created by the Late Old English vowel lengthening and shortening processes, which are the germ of a subset of the Modern English irregular verbs (type keep : kept). The underlying form of the past tense and past participle morpheme originally had not contained a vowel but was preceded by the stem-formative, but the re-interpretation postulated in (11) above created a new underlying form with a vowel, /ed/ and /od/. While /od/ was stable, /ed/ allowed deletion of the vowel in certain, originally predictable environments. This deletion process was gradually extended and became phonologically unpredictable. At the same time, /ed/ and /od/ merged in /әd/, which elevated
9 Linguistic Levels: Morphology
143
vowel-deletion to a class distinctive feature. We now have two classes: one that deletes the vowel, and one that does not. In the course of Middle English, vowel deletion was gradually generalized, at first controlled by stylistic and rhythmic factors, until it became the rule except after base-final alveolar stops. At this stage, the vowel deletion process must have been re-interpreted as a vowel-insertion process (rule inversion). There is one further important development, viz. the loss of the infinitive ending, which is already attested in the North at a relatively early stage, and gradually spreads to the South. This introduced word-based morphology also to the verb morphology, whereas it had evolved in noun morphology already in pre-Old English.
4.3 Germanic to Old English: nominal system Let me now turn to nominal inflection. A typical reconstructed Indo-European paradigm, e.g. the masculine o-stems, would have the structure shown in Table 9.2. Table 9.2: Reconstructed IE masculine o-stem (= Grmc. a-stem) paradigm of ‘day’ root SG
NOM
* h ´ hw
ACC
* h ´ hw
NOM
d og d og * h d oghw * h d oghw * h ´ hw d og *dhoghw
ACC
* h ´ hw
GEN
* h
GEN DAT PL
stem-formative
inflection
o o e´/o´ Ø/o´ː oː o´ː o o´ː/(?e´ː) o
s m so ı´/i s(es) s(es) ns m mis
d og d oghw * h ´ hw d og
DAT
Again, the stem-formatives had both derivational and inflectional function and exhibited allomorphic ablaut alternations. At this stage, the stem-formatives had a consistent, predictable exponent. In Germanic, this variable system was replaced by non-variable initial stress, which destroyed the phonological conditioning of ablaut and led to a gradual weakening and loss of inflectional syllables. This resulted in the reconstructed Germanic paradigm shown in Table 9.3 based on Voyles (1982). Table 9.3: Reconstructed Grmc. a-stem (= IE o-stem) paradigm of ‘day’ SG
NOM
*
ACC
*
GEN DAT
da´g da´g * ´g da * ´g da
+ + + +
az a es eː
PL
NOM
*
ACC
*
GEN DAT
da´g da´g * ´ dag * ´ dag
+ + + +
oːs / da´g + oːz a˜ o˜ː amz
Here it is no longer plausible to segment a stem-formative separate from the inflectional endings using conventional methods of morphological analysis, i.e. the original morphological structure had been destroyed. In forms such as SG ACC *dag-a, SG DAT *dag-eː, PL ACC *dag-a ˜ , PL GEN *dag-o˜, the original case/number endings had been lost, and in the interest of paradigm symmetry it can be assumed that the
144
II Linguistic Levels stem-formative was re-interpreted as a new case/number ending, thereby losing its original function. In the other forms, there is also no consistent representation of the old stem-formative. This resulted in a new set of inflectional endings with a concomitant shift from a ternary root + stem-formative + inflection to a binary stem + inflection structure. Further restructuring took place in the transition to Old English, as seen in the paradigm given in Table 9.4. Table 9.4: OE a-stem SG
NOM ACC GEN DAT
dæg dæg dæg + es dæg + e
PL
NOM ACC GEN DAT
dag dag dag dag
+ + + +
as as a um
With these masculine nouns and the corresponding neuters, e.g. scip, the nominative/ accusative singular had lost its exponent, resulting in a base form which could function as a word without any additional morphological material. This part of nominal inflection thus had become word-based. Since these paradigms contained the majority of nouns, they attracted more and more nouns from other paradigms, e.g. from the original -ja- and -i-stems, where a form like end-e came to be re-interpreted as ende, with -e losing its inflectional function as nominative singular and becoming part of the unmarked base form, parallel to cyning. At this stage, however, morphology was still inflectional: the two inflectional categories of number and case were fused in one exponent.
4.4 Post Old English At the end of the Old English period a number of new developments happened. These have primarily to do with the merger of unstressed vowels in schwa, i.e. luf-u > luf-e /lufə/ (original “-o¯-stem”), gum-a >gum-e /ɡumə/ (original “-n-stem”), etc., which had still been stem-based with -u and -a functioning as nominative singulars. This development makes these forms look like the continuation of the original strong -jaand -i-masculines, whose (originally inflectional) ending -e > /ə/ had already been reinterpreted as part of the base. These originally stem-based forms underwent the same reinterpretation and were integrated into the emerging generalized word-based noun morphology. Another development concerned the other inflectional forms, especially the singular genitive and dative as well as the plural nominative and accusative of these stem-based nouns. With the reduction of the final vowels to /ə/ these forms were now identical with the unmarked singular nominative, i.e. they had become endingless. It is therefore not surprising that the salient singular -s-genitive was gradually transferred to these nouns and also to other inflectional classes. At the same time, these nouns also adopted -s-forms in the plural nominative/accusative. This suggests that the ending -es had come to be reinterpreted as a general marker of the category plural only, without any additional case function. The same apparently happened to the ending -en characterizing the weak masculines of the type gum-a(n). This brings about a crucial typological realignment: the categories of number and case became dissociated,
9 Linguistic Levels: Morphology because otherwise the analogical transfer of a case ending in the singular and of a plural ending independent of a case function (plural nominative/accusative) cannot be explained. This means that already at the end of the Old English period noun inflection had become completely word-based. The gradual marginalization of the dative (both singular and plural) must have started at this period as well. It had become a basically preposition-controlled ending, whose presence or absence was more and more governed by metrical-stylistic requirements (cf. Lass 1992: 110). This indicates that its actual grammatical function had been lost at this stage (or had at least been considerably weakened), cf. comparable 19th century Ger. dem Hund-e > dem Hund. The Middle English period is characterized by the gradual generalization of the -splural, since eventually most of the -n-plurals (exceptions such as ox-en, childr-en, brethr-en are the last remnants today) adopted this plural form. Similarly, the -s-genitive singular came to be generalized. The genitive plural poses a problem, because there had never been an -s-genitive in the plural, but the irregular plural nouns – as well as the group genitive – adopted this form. For a possible explanation, cf. Kastovsky (2008). The only other remarkable restructuring between Middle English and Modern English (probably in the transition from Middle English to Early Modern English) concerns the relationship between the regular allomorphs of the plural and genitive morphemes. Originally, i.e. until Middle English, their underlying representation contained a vowel, probably /e/ or /ə/, which was raised to /ɪ/ in certain varieties. During the Middle English period, this vowel could be deleted for rhythmic reasons, so that its presence or absence became optional, except before a coronal sibilant. Eventually, this vowel deletion became the rule, just as with verbs. This generalization of the vowel deletion led to rule-inversion: the vowelless allomorph came to be interpreted as underlying, and vowel-insertion replaced vowel-deletion, controlled by phonotactic requirements. In Old English, adjectives agreed with the nominal head with regard to number, gender and case. Moreover, there were two types of inflection, so-called “weak” inflection after demonstratives, and “strong” inflection without a determiner, which was a Germanic innovation. But adjectival inflection became unstable towards the end of the Old English period and was lost in Middle English, clearly as a result of the general breakdown of NP inflection.
4.5 Word-formation Let me conclude with a few remarks on word-formation. One striking feature is the almost total loss of ablaut nouns and adjectives related to strong verbs with corresponding ablaut alternations of the type write : writ, sing : song. These had formed part of the Old English core vocabulary, as they still do in German. In English, however, they were gradually replaced by base-invariant derivatives, since native word-formation – like regular inflection – adopted base invariancy as its basic principle. The massive borrowing of Romance and Latin vocabulary in Middle and Early Modern English, however, reintroduced base alternations at a new non-native level. Another remarkable phenomenon is the demise of most Old English verbal prefixes, many of which had become semantically opaque already in Old English (cf. Hiltunen
145
146
II Linguistic Levels 1983). They were gradually replaced by non-native patterns, which dominate much more in prefixation than in suffixation. Finally, the rise of a new compound type, viz. washing machine, swimming pool in Middle English should be mentioned. This has to be seen in connection with the rise of the new -ing-participle/gerund, which replaced the old -ende-participle. But, as many other Middle English developments in this area, this still needs more detailed investigation (cf. Kastovsky 2007).
5 References Bauer, Laurie. 1983. English Word-Formation. Cambridge: Cambridge University Press. Bauer, Laurie. 1992. Introducing Linguistic Morphology. Edinburgh: Edinburgh University Press. Biese, Y. M. 1941. Origin and Development of Conversions in English. Helsinki: Finnish Academy of Sciences. Blake, Norman (ed.). 1992. The Cambridge History of the English Language. Vol. II. 1066–1476. Cambridge: Cambridge University Press. Brunner, Karl. 1960–62. Die englische Sprache. Ihre geschichtliche Entwicklung. 2 Bd. 2. Aufl. Tu¨bingen: Niemeyer. Clackson, James. 2007. Indo-European Linguistics. An Introduction. Cambridge: Cambridge University Press. Giegerich, Heinz J. 1999. Lexical Strata in English: Morphological Causes, Phonological Effects. Cambridge: Cambridge University Press. Hiltunen, Risto. 1983. The Decline of the Prefixes and the Beginnings of the English Phrasal Verb. The Evidence from Some Old and Early Middle English Texts. Turku: Turun Yliopisto. Hockett, Charles. 1958. A Course in Modern Linguistics. New York: MacMillan. Hogg, Richard M. (ed.). 1992. The Cambridge History of the English Language. Vol. I. The beginnings to 1066. Cambridge: Cambridge University Press. Hogg, Richard and David Denison (eds.). 2006. A History of the English Language. Cambridge: Cambridge University Press. Huddleston, Rodney and Geoffrey K. Pullum. 2002. The Cambridge Grammar of the English Language. Cambridge: Cambridge University Press. Kastovsky, Dieter. 1980. Zero in morphology. A means of making up for phonological losses? In: Jacek Fisiak (ed.), Historical Morphology, 213–250. The Hague: Mouton. Kastovsky, Dieter. 1996. Verbal derivation in English: A historical survey. Or: Much ado about nothing. In: Derek Britton (ed.), English Historical Linguistics 1994, 93–117. Amsterdam/ Philadelphia: John Benjamins. Kastovsky, Dieter. 1997. Morphological classification in English historical linguistics: The interplay of diachrony, synchrony and morphological theory. In: Terttu Nevalainen and Leena KahlasTarkka (eds.), To Explain the Present. Studies in the Changing English Language in Honour of Matti Rissanen, 63–75. Helsinki: Socie´te´ Ne´ophilologique de Helsinki. Kastovsky, Dieter. 1999. Inflectional classes, morphological restructuring, and the dissolution of Old English grammatical gender. In: Barbara Unterbeck and Matti Rissanen (eds.), Gender in Grammar and Cognition. Vol. 2. Manifestations of Gender, 709–727. Berlin/New York: Mouton de Gruyter. Kastovsky, Dieter. 2006a. Historical morphology from a typological point of view: Examples from English. In: Terttu Nevalainen (ed.), Types of Variation: Diachronic, Dialectal and Typological Interfaces, 53–80. Amsterdam/Philadelphia: John Benjamins. Kastovsky, Dieter. 2006b. Vocabulary. In: Hogg and Dennison (eds.), 199–270. Kastovsky, Dieter. 2006c. Typological changes in derivational morphology. In: Ans van Kemenade and Bettelou Los (eds.), The Handbook of the History of English, 151–176. Oxford: Blackwell.
9 Linguistic Levels: Morphology
147
Kastovsky, Dieter. 2007. Middle English word-formation: A list of desiderata. In: Gabriella Mazzon (ed.), Studies in Middle English. Forms and Meanings, 41–56. Frankfurt: Peter Lang. Kastovsky, Dieter. 2008. The genesis of the Modern English genitive plural: Structural and phonostylistic factors. In: Jo´zef Andor, Be´la Hollo´sy, Tibor Laczko´, and Pe´ter Pelyva´s (eds.), When Grammar Minds Language and Literature. Festschrift for Prof. Be´la Korponay on the Occasion of his 80th Birthday, 263–273. Debrecen: Institute of English and American Studies, University of Debrecen. Kastovsky, Dieter. 2009a. Diachronic aspects. In: Rochelle Lieber and Pavol Sˇtekauer (eds.), The Oxford Handbook of Compounding, 323–340. Oxford: Oxford University Press. Kastovsky, Dieter. 2009b. English word-formation, combining forms and neo-classical compounds: A reassessment. In: Current Issues in Unity and Diversity of Languages. Selected Papers from the CILT 18, Held at Korea University in Seoul on July 21–26, 2008, 724–734. Seoul: Korea University. Lass, Roger. 1990. How to do things with junk: Exaptation in language evolution. Journal of Linguistics 26: 79–102. Lieber, Rochelle. 1983. Argument linking and compounds in English. Linguistic Inquiry 14: 251–286. Lyons, John. 1977. Semantics. 2 vols. Cambridge: Cambridge University Press. Marchand, Hans. 1969. The Categories and Types of Present-Day English Word-Formation. 2nd rev. edn. Mu¨nchen: Beck. Matthews, Peter. 1974. Morphology. Cambridge: Cambridge University Press. [2nd edn. 1991]. Minkova, Donka. 2008a. Prefixation and stress in Old English. Word Structure 1: 21–52. Minkova, Donka. 2008b. Continuity or re-invention in functional stress-shifting. Paper presented at the 15th International Conference on English Historical Linguistics, Munich, August 24–30, 2008. Prc´ic´, Tvrtko. 2005. Prefixes vs initial combining forms in English: A lexicographic perspective. International Journal of Lexicography 18: 313–334. Prc´ic´, Tvrtko. 2007. Headhood of suffixes and final combining forms in English word formation. Acta Linguistica Hungarica 54: 381–392. Schlegel, August Wilhelm. 1818. Observations sur le langage et la litte´rature provenc¸ales. Paris. ¨ ber die Sprache und Weisheit der Indier: Ein Beitrag zur Begru¨ndung Schlegel, Friedrich. 1808. U der Altertumskunde. Heidelberg: Winter. Tops, Guy A. J. 1974. The Origin of the Germanic Dental Preterit. A Critical Research History since 1912. Leiden: Brill. ¨ ber die Verschiedenheit des menschlichen Sprachbaues. Bonn. von Humboldt, Wilhelm. 1827–29. U Voyles, Joseph B. 1992. Early Germanic Grammar. Pre-, Proto- and Post-Germanic Languages. San Diego: Academic Press. Wald, Benji and Lawrence Besserman. 2002. The emergence of the verb-verb compound in twentieth century English and twentieth century linguistics. In: Donka Minkova and Robert Stockwell (eds.), Studies in the History of the English Language. A Millennial Perspective, 417–447. Berlin/New York: Mouton de Gruyter. Wełna, Jerzy. 1996. English Historical Morphology. Warszawa: Wydawnictwa Uniwersytetu Warszawskiego.
Dieter Kastovsky, Vienna (Austria)
148
II Linguistic Levels
10 Linguistic Levels: Syntax 1 2 3 4 5
Introduction: the syntactic history of English and English historical syntax The syntactic history of English English historical syntax Summary References
Abstract Syntactic change in English can be considered in a number of ways, depending on whether the aim of the research is to find out more about the ways in which English has changed over time, or whether it is to find out more about general constraints on change in the syntactic component of human language by using evidence from diachronic varieties of English. Adapting a proposal from Honeybone (2008), on phonological change in English, this article is concerned with both the syntactic history of English and English historical syntax, and discusses evidence from a range of varieties of the language (both contemporary and historical). Changes discussed in the article include the development of word order patterns, the evolution of auxiliaries, and the category “subject” in English. Both formal and functional theories of language change are considered, along with some discussion of the roles played by borrowing, reanalysis, and analogy in shaping change in English syntax.
1 Introduction: the syntactic history of English and English historical syntax This section presents an overview of the ways in which the syntax of English has changed. Other contributions to this volume consider the syntax of particular periods, so this is a rather global approach to syntactic change. The overview is divided into two parts, the syntactic history of English (Section 2) and English historical syntax (Section 3), following a similar distinction regarding the evolution of English phonology proposed by Honeybone (2008). To illustrate this distinction, consider the development of the tense auxiliary do in the history of the language (see further Warner, Chapter 47). There are a number of questions one might ask about this particular change, including: a. based on the textual evidence available to us, what do we know of the regional and textual provenance of this change? b. to what extent do current varieties of English display similar or divergent patterns with respect to the use of do-support? c. how can we relate the development of do-support to other changes affecting auxiliary verbs in English, such as the modals and the aspect and voice auxiliaries? d. is this change typologically odd? e. what does this change tell us about properties of verbs and their dependents? f. how does this affect our understanding of constraints on syntactic change, or of more general issues such as reanalysis and analogy/extension? Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 148–164
10 Linguistic Levels: Syntax The first three questions relate to the syntactic history of English, since answers to such questions will tell us something about a particular development in a particular set of varieties, based on a particular set of evidence. By contrast, the last three questions relate to English historical syntax, since answers to such questions will tell us something about general properties of human language undergoing change, comparing data from (varieties of) English with other varieties as evidence for universal tendencies. It is important to keep the syntactic history of English distinct from English historical syntax, because different questions need to be asked in each case. The importance of the distinction does not lie in a difference between description and explanation: it is wrong to say that work in the syntactic history of English does not concern itself with explaining patterns of change, and equally wrong to say that matters of description are irrelevant to those working on English historical syntax. Both approaches require the appropriate use of appropriate data, and both require “theorizing”, but not necessarily in the same way. In a discussion of contemporary dialect variation and theoretical syntax, Adger and Trousdale (2007: 261) note that dialect syntax in its sociolinguistic context provides a challenge for syntactic theory because it “raises important questions regarding what the theory is actually modelling”: and the same holds true for English historical syntax. Furthermore, although the syntactic history of English and English historical syntax are distinct, they are related. This can be exemplified by situations of language contact. An accurate account of the historical syntax of English will describe the many ways in which contact with other languages (from the Celtic substratum which has been said to have had a significant effect on the evolution of British English [see various papers in Filppula et al. 2008], to the new Asian varieties which have emerged in the later modern period); but contact linguistics more generally interfaces with linguistic theory, so these developments also have a role to play in English historical linguistics (not least in the way in which phonological, morphological, and syntactic changes are themselves related, and even accelerated, in contact situations). The various inputs to a child acquiring a particular variety may differ significantly depending on the degree of imperfect second language learning which may characterize the adult’s output. Exogenous changes, brought about by borrowing from one variety into another, serve as part of the input system to subsequent generations acquiring the language (who may have no knowledge that a particular form is the consequence of borrowing), which may prompt a series of endogenous changes. However one wishes to describe this phenomenon in theoretical terms – parameter resetting, constraint reranking, constructional entrenchment, and so on – the relation between the syntactic history of English and English historical syntax is at the heart of much of the new research in this field of linguistic enquiry. My reason for detailing this distinction at some length is because much research remains to be done, despite advances made in both fields. Some work which (in part or in whole) addresses grammatical change in English is more clearly aligned with the syntactic history of English (e.g. Mitchell 1985); similarly a lot of work on historical syntax uses English as a source of data, an understandable decision given the wealth of historical material available for analysis (e.g. Lightfoot 1999); the vast majority of research includes some aspects of both (e.g. the various syntax chapters in the Cambridge History of the English Language [Hogg (ed.) 1992–2001]). However, there is still much in the domain of the syntactic history of English which remains to be uncovered; similarly, work in English historical syntax is also rapidly evolving. Some, perhaps
149
150
II Linguistic Levels most, research into English historical syntax follows mainstream work on Chomskyan generative syntax, though there is increasing work within Optimality Theory, and construction grammars of various kinds, as discussed in Section 3 below. Furthermore, some research counts as an excellent representation of both kinds of historical work, in which we discover more about both the historical development of English syntax, and the interface of these developments with more general issues in syntactic (and linguistic) theory (e.g. Warner 1993 and Allen 1995). The remainder of this introduction is concerned with some other issues of relevance to syntactic change in English. Some of these issues concern categorizations of various kinds. First, there are the rather problematic issues of what we mean by terms such as “the English language” or an “Old English dialect” when considering the changes that have occurred. Who counts as a native speaker of contemporary English, and at what arbitrary point in time do we separate Old English (OE) from Middle English (ME), for instance? Second, what evidence do we use to document the changes we observe, and how coherent are these (spoken or written) text types as instances of a single category? Comparing a Kentish charter from the 9th century with a contemporary text message written in Singlish may well indicate some of the ways in which English syntax has changed in the course of over a millennium, but we need to bear in mind not simply diachronic change, but also provenance, text type, intended communicative function, and the like. Matters such as these are perhaps more directly relevant to the syntactic history of English than to English historical syntax, but recent work on the earlier syntax of English, using the various parsed corpora, has associated different dialect patterns with both endogenous constraints and instances of language contact. A good example concerns the effects of Viking invasions on the more northerly dialects of Old English and Middle English. In terms of the particular history of English, we can witness not only place name evidence of settlements and the like, but also the structural consequences of prolonged contact with Old Norse, in the more rapid loss of inflections witnessed in texts from northern parts of England, in comparison with texts with a more southerly provenance. The syntactic consequences of this contact are far reaching, including variation in the position of finite verbs (the V2 constraint), with northern varieties of Middle English as a COMPLEMENTIZER PHRASE-V2 type, and southern varieties as INFLECTION PHRASE-V2 (Kroch et al. 2000), and, following from that, the identification of the functional projection under CP in Old English as AgrSP (Subject Agreement Phrase) (Haeberli 2000). This snapshot is intended to illustrate how it is possible to connect, via a series of steps, particular features of the syntactic history of English to highly theory-specific accounts in English historical syntax. While it is helpful to separate the syntactic history of English from English historical syntax, it is also useful to see how the two relate to one another. Very little mention is made in what follows of the methodologies adopted in accounts of syntactic change, from the primarily philological to the primarily theoretical; equally, I have decided for reasons of space not to discuss many of the ways in which the use of computerized corpora has revolutionized how work on grammatical change in English is conducted. However, both of these issues are addressed briefly in the conclusion. For further discussion of good practice in historical syntax research, see Fischer (2007: 11–52). In an overview chapter such as this, it is possible only to deal with a small subset of syntactic changes. There are three main changes I concentrate on
10 Linguistic Levels: Syntax (OV/VO word order, the loss of impersonals and the establishment of the category “modal”), linking these where possible to other changes in the system, and showing how various theoretical issues can be addressed by considering these particular changes.
2 The syntactic history of English The syntactic history of English, based on the extant materials available to us, reveals some issues of continuity throughout the history of language, e.g. the availability of preand post-modification of nouns in both Old English and Present-day English (PDE), and some aspects of change, e.g. the use of the sequence [mə] with the first person singular pronoun to indicate futurity in some non-standard varieties of English, as in I’m a get a drink. Before I go on to discuss some particular features of linguistic change, it is important to consider some aspects of the way in which the story of English is told in many historical accounts. Typically, the story is that of the evolution of forms which constitute the standard variety – explanations are given to work out why it is that the most geographically generalized and conventionalized of forms have come to take the shape that they have. There are a number of reasons for this. For instance, much work in syntactic theory has used English as a data source, and particularly, has used the idiolect of the researcher as a way of judging grammaticality, which typically reflects the middleclass, educated variety of English that typifies the standard. This is of particular relevance to English historical syntax, but still features in accounts of the history of a particular language. By way of example, we can consider the history of word order in English.
2.1 Word order changes Most research on word order in clauses has been concerned with three particular phenomena, the V2 constraint, verb raising, and OV word order, illustrated by (1), (2), and (3) respectively: (1)
On his dagum sende Gregorius us fulluht In his day-DAT.PL send-3P.SG.PAST Gregory 1P.PL.DAT baptism ‘In his time, Gregory sent us Christianity’ (ChronA2 18.565.1; Haeberli 2002: 88)
(2)
locige ic buton to ðæm eaðmodum To hwæm To who-DAT look-1P.SG.PRES I except to the-DAT humble-DAT.PL ‘To whom do I look except to the humble?’ (CP 41.299.18; Fischer et al. 2000: 67)
(3)
He ne mæg his agne aberan He NEG can his own support-INF ‘He cannot support his own’ (CP 7.53.1; Moerenhout and van der Wurff 2005: 85)
The loss of V2, OV word order and V-to-T raising has been central in much work in English historical syntax (for useful summaries within a principles and parameters model, see Fischer et al. 2000). The first two changes are typically used to explain why English word order has become more “fixed”. Exceptions to the rule in contemporary
151
152
II Linguistic Levels standard English are noted (e.g. in clauses beginning with negative adverbs, as in Never have I heard such rubbish), but typically, the story ends there. However, word order in “English” continues to evolve. The syntactic history of English is a history of contact, and in some contact varieties we see a continuation of variation in terms of V2 and verb raising in (4) and (5): (4)
if you want we can go earlier because # at four thirty starts the quiz (GermanEnglish bilingual; Eppler 1999: 303)
(5)
What he has eaten? (Indian vernacular English; Bhatt 2000: 74)
Do we include such patterns as part of the syntactic history of English? To a large extent the answer to that question depends on the degree of conventionalization in different communities. Example (4) might be a “one-off” case of interference between the competing grammars of a single bilingual (though as we will see in Section 3, the notion of competing grammars among monolinguals has been highly influential in theorizing earlier stages of word order variation and change), thus an innovation rather than a change (since the latter requires spread). But (5) is rather well established as a feature of a particular variety of English spoken by a substantial number of people. Because so much of the syntactic history of English is concerned with the tracing of what is now the standard for inner circle varieties of the language, many of the “big stories” in the syntactic history tend to come to the same conclusion, by relating what has happened to give us the contemporary standard forms.
2.2 The English modals There are also a number of questions regarding the extent of continuity in the history of the language. A useful example here comes from another of the more widely discussed changes, the story of the English modals (again a topic of relevance to both the syntactic history of English, and English historical syntax: for a useful overview of both the relevant data and different theoretical accounts, see Denison 1993: 292–339). The story of the modals is one of divergence, in which a series of (formally) slightly anomalous verbs became (formally and functionally) even more anomalous over time. Although part of a set of preterit-presents in Old English, the ancestors of modern modals had more verblike formal properties than the Present-day English modals do, such as the ability to take an object complement and to appear in non-finite form, with these properties gradually being lost over time. Allied to the formal development is a functional change, such that these verbs come to encode aspects of the speaker’s viewpoint, whether that be the intention to lay down some sort of obligation on the hearer (in deontic modality) or the assessment of the veracity of the proposition encoded elsewhere in the clause (in epistemic modality), functions which had previously been coded by subjunctive inflectional endings on verbs. The story of the modals does not involve an orderly transition from lexical verb to auxiliary. This is shown most clearly by the findings of Warner (1993), who emphasizes the apparent messiness of the development. For example, instead of developing more auxiliary-like properties in the transition from Old English to Middle English, the modals in some ways become more verb like (e.g. OE *sculan and cunnan develop new
10 Linguistic Levels: Syntax present tense forms sculeþ and cunneþ), and different dialects of Middle English show changes at different times and in different ways. What Warner’s study suggests is that the category of auxiliary (and the subcategory of modals) becomes strengthened in the history of English (see also Hudson 1997), by showing bonds between the various forms that were to emerge as auxiliaries (e.g. negative contraction in nolde ‘would not’, næs ‘was not’, nabbe ‘have not’). Warner argues that the key development is that the auxiliary category becomes more well-defined (a more basic level category), in part in terms of its interaction with other phenomena (e.g. the appearance of tag questions, negative clitics, and the position of unstressed adverbs). Towards the end of Middle English (roughly at the turn of the 15th century), the differentiation between auxiliaries and verbs becomes sharper. A further issue regarding continuity concerns the existence of “double modal” constructions in some contemporary varieties of English, as illustrated by (6) and (7): (6)
Could you might possibly use a teller machine? (Southern US English; Mishoe and Montgomery 1994: 11)
(7)
Oh no, they’re double-glazed. They wouldn’t could [break] (Tyneside English; McDonald 1981)
Such forms are well-attested in earlier varieties of English, as we see in (8): (8)
& hwu muge we þone weig cunnen and how may we the-MASC.SG.ACC way can-INF ‘And how can we know the way?’ ( Jn [Warn 30]; Fischer and van der Wurff 2006: 147)
and there is some debate regarding the extent to which the contemporary forms should be seen as continuations from earlier stages of the language, or as independent developments. Of relevance here is the fact that the syntax and semantics of modal combinations in those varieties of British English which allow them (i.e. dialects of north-eastern England and central Scotland) are rather different from those in the southern United States dialects. So it may be that we have continuation in the case of the British dialects, but independent development in the American ones (see further Nagle 1994).
2.3 Subjects and the impersonal construction My final example to illustrate the centrality of variation in the syntactic history of English concerns the category of “subject”. While it is certainly the case that there are some instances of “subjectless” clauses in early English: (9)
norþan sniwde from-north snow-3P.SG.PAST ‘It snowed from the north’ (Seafarer 31)
153
154
II Linguistic Levels such unambiguous examples are restricted either to a particular lexical set (weather verbs) (9) or to a particular kind of information packaging (impersonal passives) (10): forðy to ungemetlice (10) ðætte that therefore too greatly geliðod ðæm the-DAT let-off-PAST.PRTC ‘that therefore it must be let off too greatly et al. 2000: 39)
ne sie not be-3P.SG.PRES.SUBJ scyldgan guilty-DAT to the guilty’ (CP 20.149.24; Fischer
In cases of co-ordination such as (11): (11) and him and 3P.SG.MASC.DAT and him and 3P.SG.MASC.DAT ‘and angels came to 2000: 39)
comon englas to, come-PL.PAST angels-PL.NOM/ACC to ðenodon served him, and served him’ (ÆCHom I, 11.174.17; Fischer et al.
the absence of the subject in the second clause is predictable from the discourse context, and such VP coordination is equally common in Present-day English. More crucial are the well-known impersonal constructions such as in (12): mannes (12) him ofhreow þæ s 3SG.MASC.DAT rue-3P.SG.PAST the-GEN man-GEN ‘He pitied the man’ (ÆCHom I, 8.192.16; Denison 1993: 85) in which neither argument is marked as nominative. Such constructions have been thoroughly discussed in the literature (most comprehensively by Allen 1995), and highlight the problematic notion of “subject” as a category for earlier English. One widely held view is that the loss of impersonal constructions is a consequence of case loss (see e.g. Lightfoot 1991), where assignment of inherent/lexical (dative) case to experiencers prevents structural case assignment. The question remains, however, as to whether dative or genitive noun phrases in such constructions constitute subjects. Evidence from raising and co-ordination, as in (13) and (14) respectively:
ælmihtigum (13) him sceal sceamian ætforan gode 3P.SG.MASC.DAT shall shame-INF before God-DAT almighty-DAT ‘he shall be ashamed before God almighty’ (ÆLS [Ash Wednesday] 12.169 Visser 1970: 23) (14) ac gode ne licode na heora geleafleast … but god-DAT NEG please-3P.SG.PAST NEG 3P.PL.GEN faithlessness … ac asende him to fyr of heofnum but send-3P.SG.PAST 3P.PL.DAT to fire-ACC from heavens-DAT ‘But their faithlessness did not please God … but [he] sent them fire from heaven’ (ÆCHom II, 20.644.71; Denison 1993: 89)
10 Linguistic Levels: Syntax suggests that dative noun phrases had some subject properties; however, the category of subject was not as fully grammaticalized as it is in Present-day English, which may be associated with a more general grammaticalization of the transitive construction (Trousdale 2008). Anderson (1997: 216–224) proposes a different analysis, in which the change involves a gradual coalescence of morphosyntactic and syntactic subject. The distinction between the two can be illustrated by existential clauses such as there are books on the shelf, where books is the syntactic subject, there the morphosyntactic subject (controlling concord). In Anderson’s analysis, earlier English optionally marked morphosyntactic subjecthood, though raising in (13) suggests that the dative NP is a syntactic subject. Turning to the situation in contemporary English, as Anderson (1997: 224) observes, instances such as there’s books on the shelf suggest the wide systemic spread of the coalescence of morphosyntactic and syntactic subject. But even subjectlessness has not disappeared entirely. We find it in particular (albeit restricted) kinds of written or spoken discourse, such as diary entries of the kind went home, ate dinner, fell asleep (Haegeman 1997), or in casual dialogue (A: How many people were at the party? B: Dunno, couldn’t say), where it is understood that the subject is first person singular. There are many other aspects of the syntactic history of English which I do not have the space to deal with here (for an excellent summary, see Fischer and van der Wurff 2006; for authoritative treatment of individual periods, see the various syntax chapters of the Cambridge History of the English Language [Hogg (ed.) 1992–2001]). There are also many other topics which are still awaiting detailed treatment. Some of these are highly specific. For instance, what is the precise history of the ditransitive construction where both objects are pronouns (e.g. he sent it her vs. he sent her it)? What do we know about the spread of the progressive with stative verbs (e.g. I’m loving your new look), and how widespread is this in the new Englishes? Why does þa trigger V2 more consistently than þonne in OE? What is the spread of the indirect passive in the later history of English (and why was the spread so slow)? There are also general questions relating to on-going changes: for instance, what features characterise the syntax of emergent contact varieties involving English? Answers to these and other questions will clarify even further the various features of change in the syntactic history of English.
3 English historical syntax In this section, I look at some of the ways in which data from the history of English has been used to explore more general issues in syntactic change. I will use some of the changes discussed in Section 2 to indicate how the syntactic history of English and English historical syntax overlap, and bring in some further data to highlight some of the other relevant issues. In discussing cross-linguistic patterns, Harris and Campbell (1995: 50) state that there are three mechanisms of syntactic change: borrowing, reanalysis, and extension. I deal with each in turn, though for reasons of space I say less about borrowing than about the other mechanisms of change, and I give generally accepted definitions of each mechanism. Borrowing, in its strictest sense, occurs when the “replication of the syntactic pattern is incorporated into the borrowing language through the influence of a host pattern found in a contact language” (Harris and Campbell 1995: 51); reanalysis involves “change in the structure of an expression or class of expressions that does not involve any immediate or intrinsic modification of its surface
155
156
II Linguistic Levels manifestation” (Langacker 1977: 58), while extension is the reverse, involving changes “in the surface manifestation of a pattern […] which does not involve immediate or intrinsic modification of its underlying structure” (Harris and Campbell 1995: 51). Harris and Campbell see extension as part of analogy, but avoid using this second term because it has a range of meanings in the literature. The critical issue is whether the analogy involves some exemplar or not, as discussed below: exemplar-based analogy may be equated with extension, but “non-exemplar based analogy” is rather different, as Kiparsky (forthc.) has argued.
3.1 Borrowing Research on borrowing in the earlier history of the language (i.e. on English spoken in Britain) has considered both borrowing from other languages and borrowing from different dialects of English. In terms of language contact, for earlier English, syntactic borrowing from the Celtic substratum (Filppula et al. 2008) and from Old Norse (McWhorter 2002) has been suggested, and the effects of long-term contact with Vikings in the north of England have been shown to be widespread (see further below). An interesting case regarding borrowing from Latin concerns the development of accusative with infinitive constructions in Middle English: Fischer (1989) suggests that Latin borrowing alone was not responsible for the change, though the existence of such constructions in Latin may have facilitated some aspects of its spread. The later history of language contact involving English has more to do with English beyond Britain, particularly in the development of African American English, and a range of pidgin and creole languages. Dialect contact has also been influential in the development of particular morphosyntactic features of both British English, and other Englishes, e.g. in the spread of do-support, and the variation between -th, -s and zero as inflections on third person singular indicative verbs (Nevalainen and Raumolin-Brunberg 2003).
3.2 Reanalysis Reanalysis and extension/analogy have been at the centre of much of the work in English historical syntax. In research on many of the changes discussed here (e.g. OV/ VO word order, loss of the impersonal construction, and the establishment of the category modal), two rather different claims have been made. One is that change is catastrophic, the other is that it is gradual. This raises the question of what we mean by syntactic change, and different kinds of grammarians are likely to give different answers to that question. I deal first with formal approaches to syntactic change, and how these approaches deal with the issue of reanalysis. Later in this section, I consider functional accounts of reanalysis in syntactic change. For many formal grammarians, what changes is the system, the set of parameters (or features on lexical items) which determine well-formedness in a particular manifestation of human language. In this approach to change, the primary focus is on acquisition, on how a child acquiring a particular language comes to set parameters in a particular way, given a particular set of inputs. Acquisition is the locus of reanalysis, as the language learner sets the parameters of his or her grammar, based on the primary linguistic data. The notion of parameters has changed as Chomskyan theory has developed, with specific consequences for this particular theory of syntax and our understanding of
10 Linguistic Levels: Syntax syntactic change more generally, and parametric change has been suggested by some (e.g. Pintzuk et al. 2000) to be reduced to the creation and combination of feature bundles in lexical items. This diachronic notion of the locus of change correlates with other Minimalist work on synchronic dialectal variation (e.g. Adger 2007 on variation in Buckie Scots). Increased use of corpora in formalist accounts of change, combined with particular discussions of dialectal differences in the implementation of changes (e.g. Kroch and Taylor 1997 on verb movement in Middle English) has rather altered some of the perceptions of the importance of E(xternalized)-language data in accounting for I(nternalized)-language changes. Formalist attempts to associate statistical patterns in corpora/E-language to changes in a particular individual, mental system/ I-language have re-evaluated some of the claims regarding the ways in which reanalysis is actualized. Nonetheless, reanalysis is the primary mechanism of change in most formal accounts, a mechanism of change which is also discussed in the more functionalist grammaticalization literature. Indeed Roberts and Roussou (2003), adopting a Minimalist theory of language structure, argue that grammaticalization is an epiphenomenon, since it is simply a particular case of parametric change involving reanalysis – the production of functional material from lexical material (primary grammaticalization) or other functional material (secondary grammaticalization) – and structural simplification. This focus on functional heads – and the notion that functional heads are the “magnets” which trigger Move (Chomsky 1995) – is also used to explain patterns of word-order variation, a linguistic feature which Meillet (1958 [1912]) also considered to be associated with grammaticalization. However, while Roberts and Roussou identify similarities between patterns of word order variation and instances of grammaticalization, they also show how the two are distinct. For instance, in their Minimalist model, all of the word order changes mentioned in the previous section – the loss of V2, the loss of verbraising/V-to-T, and the shift from OV to VO – involve “loss of movement to a higher functional position” (Roberts and Roussou 2003: 206). Loss of movement also characterizes the development of the modals (Roberts and Roussou 2003: 195), which unlike the others is an instance of grammaticalization. They therefore identify five significant differences between the two instances of loss of movement, one of which entails grammaticalization, the other of which does not, for example: a. only the grammaticalization of the modals creates a “new realization for T (T* Merge)” (Roberts and Roussou 2003: 207) b. the word order changes involve a reanalysis to a lower functional head (e.g. C(OMPLEMENTIZER)-to-T), the modals to a higher one (i.e. V-to-T) c. loss of V-to-T affects all lexical verbs, but the changes in the modals affect only a subset (but see Hudson 1997 for an alternative analysis which dispenses with V-to-T movement) d. upward reanalysis only is associated with bleaching and phonetic attrition (though this seems rather a stipulation than anything explanatory) e. upward reanalysis can be cyclical. Associated with this account of parameter resetting is the notion of grammar competition (Kroch 1989) as an explanation of language change. In grammar competition, an individual speaker is said to display patterns of variation that cannot be the product
157
158
II Linguistic Levels of operations of a single grammar. In a principles and parameters model, it is impossible for an individual speaker to have one grammar that both allows verb raising and simultaneously disallows it: the “switch” for the V-to-T parameter cannot be set at both “on” and “off” in one individual grammar. In more recent Chomskyan theory, as noted above, the focus of grammar competition has shifted to features of lexical items (and in cases of variation, how the same lexical item surfaces with apparently contradictory feature markings). A further issue associated with grammar competition is the Constant Rate Effect (Kroch 1989), a phenomenon used to link frequency with rates of change: for any change involving grammar competition, while the change may occur more frequently in one syntactic context than in another, the rate of change across different contexts remains the same. Different rates of change indicate that there is likely to be more than one change involved. Grammar competition has been used as an explanation for word order changes such as the loss of OV. In an elegant study (which also illustrates the usefulness of computerized corpora in syntactic change), Pintzuk and Taylor (2006) provide quantitative evidence that, although VO order occurs with different kinds of objects (positive, quantified, and negative) all affected by different kinds of factors (such as length and thematic role), the rates of change are different, and so cannot be explained simply in terms of grammar competition; rather, there is both (a) grammar competition between head initial and head final VPs and (b) addition stylistic motivations for objectmovement. Crucially, the corpus data provide no evidence of negative objects postposing with OV grammars, and little evidence of positive objects preposing with VO grammars. The rise of VO and the demise of OV, in this account, is due to not only the loss of the OV grammar, but also the loss of the movement rule in VO generated grammars. Given that the rate of loss of the preposing movement rule is different for different kinds of objects (i.e. positive or quantified), Pintzuk and Taylor account for the gradual nature of the actualization of the changes in the history of the language. In such an approach to change, the actualization of change may be gradual, and this gradualness may be systemic (changes may occur in a particular syntactic context before spreading to another), spatial (affecting the idiolects of one geographical area before another), stylistic (originating in a particular speech context, or register) or social (occurring in the language of a particular subgroup in the speech community before being transmitted to another subgroup), or any combination of these. However, the reanalysis itself is “abrupt and catastrophic” (Lightfoot 1999: 88), and what changes is grammar. By contrast, in some functionalist analyses of grammatical change, what changes is use. Indeed, in usage-based models (Kemmer and Barlow 2000) the relationship between form and function is something of a feedback loop, where “usage feeds into the creation of grammar, just as much as grammar determines the shape of usage” (Bybee 2006: 730); the frequency of use affects the mental representation of language, such that “grammar is the cognitive organization of one’s experience with language” (Bybee 2006: 711). In models of language structure such as these, reanalysis occurs, but it is not restricted to changes of features or resetting parameters; instead, what is involved in change is form-function reanalysis, alterations to “the form-meaning mapping in a grammatical construction” (Croft 2000: 118). This notion of reanalysis is central to Hollmann’s work on the development of the have causative in English (e.g. my boss had me work late) arising from an “affecting event” construction (e.g. he wolde haue his reign endure and last), in which Hollmann argues that the form-function
10 Linguistic Levels: Syntax reanalysis is based on alternative construals: the pragmatics of an experiential event allow for a causative interpretation where the experiencer is understood to be more powerful than the other participant in the process (Hollmann 2003: 87): compare (15) and (16): (15) I often have my boss come in when I’m sleeping (= ‘It often happens that my boss comes in when I’m sleeping’) (16) My boss often has me come in when he’s sleeping (= ‘My boss forces me to come in when he’s sleeping’) While both examples could be interpreted as either causative or as “affecting event” constructions, (15) seems more likely as an affecting event, and (16) as a causative, because of the different power relations inherent in an employer-employee relationship. Hollmann’s account shows how a cluster of constructions involving have with a different (formal) complement had a similar function to the emerging have causative in late Middle English. Reanalysis here, then, is conceived as accommodation attempts: speakers and hearers negotiating the relationship between the forms and functions of constructions while trying to maintain their conventionalized uses (see further Croft 2000: 118). Change in this sense is most likely to be gradual, because it may involve incremental changes at all parts of a construction. We can consider such an instance of gradual change in the development of the English determiner. In their discussion of the elements of the noun phrase, Fischer and van der Wurff (2006: 114) argue that there has been little change in the order of the various dependents, suggesting that there has been no new functional slot in the history of the language. This position runs counter to that of Denison (2006: 288), who suggests that, rather, parallel to the development of the category Modal in English, the “evidence for the existence of D[eterminer] is much shakier in earlier English”, and somewhat problematic even in contemporary English. Categorization – both the general cognitive process and the language-specific outcomes – has been, and continues to be, an important issue in general linguistics, but it is of particular concern in historical work, where issues of synchronic gradience at time 1, and diachronic gradualness, either temporally (i.e. between time 1 and time 2), or structurally (e.g. between a change in some aspect of meaning and a change in some aspect of form), or sociolinguistically (e.g. in the spread between groups of speakers, or registers), are of considerable importance, as noted above. Specifically in the syntactic history of English, we certainly see a strengthening of the category “determiner” over time. We see this development most sharply, perhaps, in the development of articles, the most grammaticalized of determiners. In this development, English is like many other languages, in that the definite article the has arisen from a demonstrative, while the indefinite article a(n) has arisen from the numeral “one” (OE an). In less formal contemporary English, the use of some with a singular count noun (as in Some guy called earlier) we witness a further typical grammaticalization path, through which an indefinite article emerges from and becomes layered with an earlier quantifier (OE sum). The absence of such clearly grammaticalized articles in the Old English period does not, of course, mean that speakers of the language at that time could not express definiteness; the difference was simply in
159
160
II Linguistic Levels the means of expression: strong forms of adjectives typically indicated indefiniteness, as in (17): (17) ðurh boclic-e lare through book-like teaching ‘by teaching with books’ (ÆCHom Pref. 175.68) while weak forms (often with a demonstrative, suggesting incipient grammaticalization even at the earliest stages of the language) typically indicated definiteness (18): (18) se
frumsceapen-a mann first-created-NOM.SG.MASC man ‘the first man’ (ÆCHom I 7.240.250)
DEM.NOM.SG.MASC
In other words, what has remained stable is the capacity to mark particular “things” as (in)definite; what has changed is the means by which this is marked. Put another way, speakers of English have always had some way of grounding nominals, but over time they have chosen to do this using different linguistic strategies: the function is constant, the form changes. This observation from history is of relevance not just to our understanding of the categories of Present-day English, but more generally for our understanding of the nature of categories. Particularly, it brings into question whether categories may be determined by syntax alone (see Anderson 1997 and Aarts 2007 for different views on this), and it helps us to understand the relationship between gradience and gradualness in grammatical change (Traugott and Trousdale 2010). The crystallization of articles forms part of a more general development in the gradual evolution of the category “determiner” in Denison’s account; many of these crystallizations may be considered as instances of grammaticalization, including the deictification of post-determiners such as various and several (Davidse and Breban 2006).
3.3 Analogy Analogy has been an equally debated concept in syntactic change. Both reanalysis and analogy have been widely debated in the grammaticalization literature (see Hopper and Traugott 2003 for a useful summary); recently Fischer (2007) has called for a greater focus on the role of analogy in grammaticalization (arguing, for example, that in early Modern English be going to joins the token set of [AUX V] based on analogy with other future markers like will), while Kiparsky (forthc.) has equated non-exemplar based analogy with grammaticalization. Fischer’s model presents an interesting synthesis between iconic and indexical relations, between types and tokens (and relevant sets for each), and between the various paradigmatic and syntagmatic processes in grammaticalization; the model also raises the important issue of whether reanalysis results in a totally new structure, but see also Meillet (1958 [1912]) on analogy, grammaticalization and the creation of new forms. Kiparsky’s distinction between exemplar based and non-exemplar based analogy make very interesting predictions for unidirectionality in syntactic change. Because exemplar-based analogy is local (language-specific), there are exceptions to it, but it is nonetheless an attempt by
10 Linguistic Levels: Syntax the language user to simplify the grammar by abandoning redundancies; by contrast, the only “model” for non-exemplar analogy has to be the most general of grammatical constraints, universal grammar itself. It is this non-exemplar based analogy that is more commonly known as grammaticalization in the sense of Meillet (1958 [1912]): the creation of new forms. For Kiparsky, then, analogy and grammaticalization both instantiate types of grammar optimization. From a functionalist perspective, a rather different way of unifying reanalysis and analogy comes from grammatical constructionalization, the entrenchment of schematic constructions through a series of discrete reanalyses, motivated by pattern matching – recall the definition of reanalysis provided by Croft (2000) above, as a “change to the form-meaning mapping of a grammatical construction” – this series of discrete reanalyses give the appearance of gradual change, and can be exemplified by a series of changes in English grammar, including the development of degree modifiers (Denison 2002; Traugott 2008a) and cleft constructions (Traugott 2008b; Patten 2010).
4 Summary The one thing that is constant about the history of English syntax is variation. But questions still remain regarding the best way of capturing and modelling that variation optimally. Certainly work using computerized corpora has revolutionized what can be done in the quantitative (and qualitative) analysis of syntactic change. In relation to the history of English syntax, this has been most firmly established in work that intersects with the generative approach to diachronic syntax, as we have seen; but more recently, in cognitive linguistics, we see the use of corpora to explain patterns of collostructional variation and change in new Englishes (Mukherjee and Gries 2009) and in other languages (Hilpert 2008). This has the potential to be a very illuminating way of exploring many of the major issues in syntactic change from a cognitive, usage-based perspective. In the overall history of the syntax of the language, we see constant processes of renewal, as similar functions come to be coded in different ways. Particularly, we see renewal as part of grammaticalization, such that terms expressed by means of inflection at one stage in the language come to be expressed periphrastically, by means of the syntax, at a subsequent stage. The importance of contact on particular developments in English syntax cannot be understated. First, it played a significant role in the development of the determiner and auxiliary system, and on the establishment of particular word orders, in the earlier history of English; but second, and equally importantly, as English continues to be used in different communities around the world, speakers of the new varieties which emerge from contact with other languages conventionalise new patterns, so a subsequent wave of variation begins. Such external factors combine with system-internal changes – loss of phonological contrasts, and increased morphological syncretism – and lead to a series of restructurings. Thus histories of the syntax of English must take into account both the particular social and linguistic context in which speakers have used, and continue to use, varieties of English, and the more general patterns of syntactic change which can be witnessed in any human language. Both kinds of change require theorization and explanation, and neither can be achieved by appeal solely to local customs of use, or solely to cross-linguistic tendencies.
161
162
II Linguistic Levels Acknowledgments: I would like to thank Laurel Brinton and Elizabeth Traugott for their very helpful comments on an earlier version of this article.
5 References Aarts, Bas. 2007. Syntactic Gradience. Oxford: Oxford University Press. Adger, David. 2007. Combinatorial variability. Journal of Linguistics 42: 503–530. Adger, David and Graeme Trousdale. 2007. Variation in English syntax: Theoretical implications. English Language and Linguistics 11: 261–278. Allen, Cynthia. 1995. Case Marking and Reanalysis. Oxford: Clarendon. Anderson, John M. 1997. A Notional Theory of Syntactic Categories. Cambridge: Cambridge University Press. Bhatt, Rakesh M. 2000. Optimal expressions in Indian English. English Language and Linguistics 4: 69–95. Bybee, Joan. 2006. From usage to grammar: The mind’s response to repetition. Language 82: 711–733. Chomsky, Noam. 1995. The Minimalist Program. Cambridge, MA: MIT Press. Croft, William. 2000. Explaining Language Change. London: Longman. Davidse, Kristin and Tine Breban. 2006. Deictification: The Development of Postdeterminer Uses of Adjectives. Preprint no. 250. Department of Linguistics, KU Leuven. Denison, David. 1993. English Historical Syntax: Verbal Constructions. London: Longman. Denison, David. 2002. History of the sort of construction family. Paper presented at the International Conference on Construction Grammar 2, Helsinki, 7 September 2002. Denison, David. 2006. Category change and gradience in the determiner system. In: van Kemenade and Los (eds.), 279–304. Eppler, Eva. 1999. Word order in German English mixed discourse. UCL Working Papers in Linguistics 11: 285–308 Filppula, Markku, Juhani Klemola, and Heli Paulasto (eds.). 2008. English and Celtic in Contact. London: Routledge. Fischer, Olga. 1989. The origin and spread of the accusative and infinitive construction in English. Folia Linguistica Historica 8: 143–217. Fischer, Olga. 2007. Morphosyntactic Change: Functional and Formal Perspectives. Oxford: Oxford University Press. Fischer, Olga, Ans van Kemenade, Willem Koopman, and Wim van der Wurff. 2000. The Syntax of Early English. Cambridge: Cambridge University Press. Fischer, Olga and Wim van der Wurff. 2006. Syntax. In: Richard Hogg and David Denison (eds.), A History of the English Language, 109–198. Cambridge: Cambridge University Press. Haeberli, Eric. 2000. Adjuncts and the syntax of subjects in Old and Middle English. In: Pintzuk et al. (eds.), 109–131. Haeberli, Eric. 2002. Inflectional morphology and the loss of verb second in English. In: David Lightfoot (ed.), Syntactic Effects of Morphological Change, 88–106. Oxford: Oxford University Press. Haegeman, Liliane. 1997. Register variation, truncation, and subject omission in English and French. English Language and Linguistics 1: 233–270. Harris, Alice C. and Lyle Campbell. 1995. Historical Syntax in Cross-linguistic Perspective. Cambridge: Cambridge University Press. Hilpert, Martin. 2008. Germanic Future Constructions: A Usage-based Approach to Language Change. Amsterdam/Philadelphia: John Benjamins. Hogg, Richard M. (ed.). 1992–2001. The Cambridge History of the English Language. Cambridge: Cambridge University Press Hollmann, Willem. 2003. Synchrony and Diachrony of English Periphrastic Causatives: A Cognitive Perspective. Ph.D. dissertation, University of Manchester.
10 Linguistic Levels: Syntax Honeybone, Patrick. 2008. Historical phonology: Phonological variation and language change. Keynote seminar, Nordic Language Variation Network. University of Bergen/Fla˚m. Hopper, Paul and Elizabeth Closs Traugott. 2003. Grammaticalization. 2nd edn. Cambridge: Cambridge University Press. Hudson, Richard. 1997. The rise of auxiliary do: Verb-non-raising or category strengthening? Transactions of the Philological Society 95: 41–72. Kemmer, Suzanne and Michael Barlow. 2000. Introduction: A usage-based conception of grammar. In: Michael Barlow and Suzanne Kemmer, (eds.), Usage-based Models of Language, 7–28. Stanford: CSLI. Kiparsky, Paul. forthc. Grammaticalization as optimiziation. In: Dianne Jonas and John Whitman (eds.), Grammatical Change: Origins, Nature, Outcomes. Oxford: Oxford University Press. Kroch, Anthony. 1989. Reflexes of grammar in patterns of language change. Language Variation and Change 1: 199–244. Kroch, Anthony and Ann Taylor. 1997. Verb movement in Old and Middle English: Dialect variation and language contact. In: Ans van Kemenade and Nigel Vincent (eds.), Parameters of Morphosyntactic Change, 297–325. Cambridge: Cambridge University Press. Kroch, Anthony, Ann Taylor, and Donald Ringe. 2000. The Middle English Verb-Second constraint: A case study in language contact and language change. In: Susan C. Herring, Pieter van Reenen, and Lene Schøsler (eds.), Textual Parameters in Older Languages, 353–391. Amsterdam/ Philadelphia: John Benjamins. Langacker, Ronald W. 1977. Syntactic reanalysis. In: Charles Li (ed.), Mechanisms of Syntactic Change, 57–139. Austin: University of Texas Press. Lightfoot, David. 1991. How to Set Parameters: Arguments from Language Change. Cambridge, MA: MIT Press. Lightfoot, David. 1999. The Development of Language: Acquisition, Change, and Evolution. Oxford: Blackwell. McDonald, Christine. 1981. Variation in the Use of Modal Verbs with Special Reference to Tyneside English. Ph.D. dissertation, University of Newcastle upon Tyne. McWhorter, John. 2002. What happened to English? Diachronica 19: 217–272. Meillet, Antoine. 1958 [1912] L’evolution des formes grammaticales. Reprinted in: Linguistique Historique et Linguistique Ge´ne´rale, 130–158. Paris: Champion. Mishoe, Margaret and Michael Montgomery. 1994. The pragmatics of multiple modal variation in North and South Carolina. American Speech 69: 3–29. Mitchell, Bruce. 1985. Old English Syntax. Oxford: Clarendon. Moerenhout, Mike and Wim van der Wurff. 2005. Object-verb order in early sixteenth-century English prose: An exploratory study. English Language and Linguistics 9: 83–114. Mukherjee, Joybrato and Stefan Th. Gries. 2009. Collostructional nativisation in New Englishes: Verb-construction associations in the International Corpus of English. English World Wide 30: 27–51. Nagle, Stephen J. 1994. The English double modal conspiracy. Diachronica 11: 199–212. Nevalainen, Terttu and Helena Raumolin-Brunberg. 2003. Historical Sociolinguistics. London: Longman. Patten, Amanda L. 2010. Grammaticalization and the it-cleft construction. In: Traugott and Trousdale (eds.), 221–243. Pintzuk, Susan and Ann Taylor. 2006. The loss of OV order in the history of English. In: Ans van Kemenude and Bettelou Los (eds.), The Handbook of the History of English, 249–278. Oxford: Blackwell. Pintzuk, Susan, George Tsoulas, and Anthony Warner. 2000. Syntactic change: Theory and method. In: Pintzuk et al. (eds.), 1–22. Pintzuk, Susan, George Tsoulas, and Anthony Warner (eds.). 2000. Diachronic Syntax: Models and Mechanisms. Oxford: Oxford University Press.
163
164
II Linguistic Levels Roberts, Ian and Anna Roussou. 2003. Syntactic Change: A Minimalist Approach to Grammaticalization. Cambridge: Cambridge University Press. Traugott, Elizabeth Closs. 2008a. Grammaticalization, constructions and the incremental development of language: Suggestions from the development of degree modifiers in English. In: Regine Eckardt, Gerhard Ja¨ger, and Tonjes Veenstra (eds.), Variation, Selection, Development: Probing the Evolutionary Model of Language Change, 219–252. Berlin/New York: Mouton de Gruyter. Traugott, Elizabeth Closs. 2008b. “All that he endeavoured to prove was …”: On the emergence of grammatical constructions in dialogic contexts. In: Ruth Kempson and Robin Cooper (eds.), Language in Flux: Dialogue Coordination, Language Variation, Change and Evolution, 143– 177. London: Kings College Publications. Traugott, Elizabeth Closs and Graeme Trousdale. 2010. Gradience, gradualness and grammaticalization: how do they intersect? In: Traugott and Trousdale (eds.), 19–44. Traugott, Elizabeth Closs and Graeme Trousdale (eds.). 2010. Gradience, Gradualness and Grammaticalization. Amsterdam/Philadelphia: John Benjamins. Trousdale, Graeme. 2008. Words and constructions in grammaticalization: The end of the English impersonal construction. In: Susan Fitzmaurice and Donka Minkova (eds.), Studies in the History of the English Language IV: Empirical and Analytical Advances in the Study of English Language Change, 301–326. Berlin/New York: Mouton de Gruyter. Visser, F. Th. 1970. An Historical Syntax of the English Language. Part One: Syntactical Units with One Verb. Leiden: E. J. Brill. Warner, Anthony. 1993. English Auxiliaries: Structure and History. Cambridge: Cambridge University Press.
Graeme Trousdale, Edinburgh (UK)
11 Linguistic Levels: Semantics and lexicon 1 2 3 4 5 6 7 8 9
Introduction Cognitive semantics and metaphor Invited inferencing and conceptual metonymy Collocation and collostructional analysis Productivity of semantic changes at specific periods Differences between lexical and grammatical changes Changes in the lexicon Future prospects References
Abstract Selected topics in research on semantic change are discussed with focus on work based in Cognitive Linguistics and neo-Gricean pragmatics. Recent work on metaphor and metonymy, grammaticalization, subjectification, and collostructional analysis are highlighted and shown to provide theoretical underpinnings for some traditional taxonomies of semantic change. As the inventory of form-meaning pairs in a language, the historical lexicon reflects semantic change; it also reflects how vocabulary size has changed in English, and how borrowings have affected the typological ways in which meanings are packaged into words. Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 164–177
11 Linguistic Levels: Semantics and lexicon
1 Introduction While meaning or “semantics” has been a central concern of philosophy since early Greek times, semantic change has been a subject of investigation for little more than a hundred years. A landmark work is Bre´al (1897), in which a taxonomy of semantic changes was developed (and revised in structuralist terms by Ullmann 1964) that has in essence been repeated in most textbooks and handbooks on language change (e.g. Hock and Joseph 1996; Campbell 2004; Fortson 2003). Key concepts include change due to metaphor (e.g. tissue ‘woven cloth’ > ‘aggregation of cells in animals or plants’) and metonymy (e.g. board ‘table’ > ‘people sitting around a table, governing body’), pejoration (conceit ‘(self)concept’ > ‘overestimation of one’s qualities’) and amelioration (e.g. ME nice ‘foolish’ > ‘pleasant’), narrowing (e.g. OE mete ‘(solid) food’ > ‘food derived from animals’) and broadening (e.g. ME bridde ‘nestling’ > bird), and taboo avoidance (e.g. 16th century toilet ‘cloth’ > 17th century ‘cloth covering for dressingtable’ > 18th century ‘dressing-table’ > 19th century ‘lavatory’ (a euphemistic use) > 20th century ‘bathroom fixture’). Often several types of change affect one item, e.g. the changes to toilet involve first narrowing (restriction to a certain type of cloth), metonymy (object for object covering it, a subtype of whole for part), taboo avoidance, and then further narrowing. Such examples suggest that semantic change is haphazard and unpredictable. Central to Bre´al’s theory of semantics and semantic change was the importance not only of reference to objects in the world but also of what we now call “sense” (meaning defined in terms of linguistic relations), and of polysemy (related meanings associated with the same form). With the advent of structuralism, differences in sense came to be thought of in terms of “lexical fields” consisting of tightly-knit sets of words with similar meaning, such as terms for intellectual cleverness, colors, or kinship (for examples and references, many of them German, see Ullmann 1964: 243–253). In much of this work it was assumed that there were relatively fixed components of meaning, and that they could be organized in different ways across languages and times. In the latter part of the 20th century, semantic change came to be rethought in terms of more flexible sets of semantic properties. For example, one can think of how the term car fits into a “semantic space” devoted to vehicles like tank, plane, with respect to such factors as their constituent parts (metal, tires), shape, purpose, and how they are driven. Such factors are called “qualia” in Pustejovsky’s (1995) Generative Lexicon theory. The flexibility of categories is highlighted in work on prototypes and categories with fuzzy boundaries that change over time (see Geeraerts 1997; Grondelaers et al. 2007). Research on cognitive semantics, Gricean pragmatics, and grammaticalization showed that semantic change was more frequently replicated and “regular” than had often been assumed, and the advent of electronic corpora made fine-grained analysis of change in context possible. Meaning change can be conceptualized along two dimensions. One is “semasiological”: attention is paid to how meaning changes, while form remains reasonably constant (but subject to phonological change). The questions are, what meanings does a word have, how are they related, and how did they arise over time? Most of the changes listed in the taxonomies were thought of in terms of semasiology. The other dimension is “onomasiological”: attention is paid to relations that hold between the lexical items in semantic space and which forms come to express a certain meaning. This is the
165
166
II Linguistic Levels principle behind Buck’s (1949) dictionary of synonyms in Indo-European languages. It is also one of the principles behind the Historical Thesaurus of the Oxford English Dictionary (Kay et al. 2009). Current work on onomasiology concerns a wide-ranging set of concepts from expressions of emotion such as anger and pain (see Dı´az Vera 2002), to intelligence and stupidity (Allan 2008), adjectives of difference like distinct, several (Breban 2008), and modality, especially as expressed by semi-modals like have to, dare (to) (Krug 2000). The meaning of a word is only part of what we know about it. We also know its morphosyntactic structure, its phonology, what register it belongs to, perhaps what language it was borrowed from, etc. These complex sets of information are specified in the “lexicon”, the inventory of form-meaning pairs in a language. This chapter ends with some discussion of the lexicon from a historical perspective (see Section 7). In this chapter I focus on aspects of recent research on semantic change, mostly from a cognitive perspective.
2 Cognitive semantics and metaphor Cognitive Linguistics, as developed in the 1970s and 1980s, theorized a view of linguistic structures “not as if they were autonomous, but as reflections of general conceptual organization, categorization principles, processing mechanisms, and experiential and environmental influences” (Geeraerts and Cuyckens 2007: 3). A fundamental claim is that words do not have stable meanings. Rather, they are cues to potential meaning, or instructions to create meanings as words are used in context (Warren 1992, and especially 1999). These meanings are non-discrete and have prototypical properties, with core and peripheral readings (Geeraerts 1997). Furthermore, they may have rich polysemic structures (in contrast to monosemous views of semantics such as Relevance Theory, Sperber and Wilson 1995). In Cognitive Linguistics a sharp distinction between semantics and pragmatics is rejected. Although details of the theory vary considerably from author to author, there has been relative convergence among those who do work on historical semantics. The central research questions have concerned metaphor and metonymy, and ways in which semantic change occurs in linguistic contexts. Sweetser (1990) proposed a theory of metaphor and metaphorical change drawing on theories of embodiment (e.g. Lakoff 1987) and force-dynamics including exertion of force and blockage by barriers (e.g. Talmy 1988). Sweetser argued, for example, that a metaphor such as KNOWING IS SEEING developed in Indo-European languages from embodied perceptual capacities such as seeing, hearing, and grasping, and that mapping from the socio-physical world of embodiment to the abstract epistemic one of reasoning accounted for the directionality of such cross-linguistically attested meaning changes as PIE *weid ‘see’ > wit, and Gk. oı˜da perfective of eidon ‘to see’ > idea, or must ‘be required’ (compelled by socio-physical force; deontic) > ‘can be inferred’ (compelled by reasoning; epistemic). One of Sweetser’s hypotheses was that since meaning change is not random, there must be constraints on mapping from one domain to another. These constraints involve “certain abstract and topological aspects of semantic structure, which we have termed image-schematic structure, […] which must be preserved across metaphorical mappings” (Sweetser 1990: 59, italics original). Such image-schematic structures may involve barriers, e.g. may of permission represents a potential barrier that is not yet in place whether in the socio-physical domain (permission), or the domain of reasoning (possibility).
11 Linguistic Levels: Semantics and lexicon Much of the initial work on metaphor was synchronic and lexical. Particularly influential in this respect was Reddy’s (1993 [1979]) study of the conduit metaphors used for language, which accounts for extensive use of expressions concerned with language conceptualized as a physical pipe-line, e.g. put something into words, the letter contains many typos. In historical work, however, metaphor came to be noted especially in studies of grammaticalization, understood at the time as the use of lexical expressions to serve a grammatical function (see Hoffmann and Brems, Volume 2, Chapter 99). Crosslinguistically, temporal expressions derive from spatial ones (Heine et al. 1991; Haspelmath 1997), e.g. after ‘behind+COMPR’; (cf. aft of a ship); prepositions and adverbs from body parts (e.g. behind, ahead), etc. In this work, the two dimensions of semasiology and onomasiology were fruitfully combined. Heine and Kuteva’s (2002) World Lexicon of Grammaticalization provides an appendix with cross-linguistically attested source > target meaning change; this is a semasiological approach tracking changes of meaning (e.g. BODY > reflexive, cf. self ). A second appendix outlines target < source meaning change, an onomasiological approach tracking where expressions of a particular meaning derive from (e.g. FRONT < FACE, HEAD, MOUTH). Metaphor operates on the dimension of choice from among related meanings, and therefore of analogy, iconicity, and paradigmaticity. On the other hand, metonymy, being associative, operates on the dimension of indexicality, linear production, and perception (Anttila 1989: 142). Metonymy and metaphor often intersect at the conceptual level, and indeed the metaphors of embodiment ultimately derive from metonymic associations. Barcelona hypothesizes that “the target and/or source must be understood or perspectivized metonymically for the metaphor to be possible” (Barcelona 2000: 31). On this view, metonymy activates mental access to another domain, e.g. the metaphor SADNESS IS DOWN derives from experiential association with the downward bodily posture that people tend to adopt when they are sad (see also Ko¨vecses and Radden 1998; Panther and Thornburg 2003).
3 Invited inferencing and conceptual metonymy Sweetser’s work engendered tremendous interest in semantic change as metaphorical change. However, challenges and alternative proposals abounded as well. For example, Bybee (2007: 978) points out that image-schema preservation cannot apply to some changes, such as the change from perfect/anterior > present in inchoative (change of state) or state verbs, since there is no plausible image-schema associated with the perfective. In a language like Island Carib, some perfective verbs are used to express present state. Bybee suggests that an expression meaning ‘It has turned ripe’ is relevant only if it is still ripe at the time of utterance. It is the meaning of “present relevance” that underlies the change. Metaphor cannot be involved here. Rather, what is in operation is the associative, metonymic process called “invited inferencing”. We may note that most of the “core” modals in English were once “preterit-presents”, i.e. state verbs with past morphology; Gk. oı˜da cited above is the perfective form; and must is a past tense form of mot- ‘be able’ (an originally preterit-present verb). In fact, when semantic changes are considered in context, not in the decontextualized format of the examples I have cited above, they may be construed rather differently. What are schematically presented as see > ‘understand’, must ‘obligation’ > ‘epistemic conclusion’, be going to ‘motion with a purpose’ > ‘future’ are abstractions over many
167
168
II Linguistic Levels centuries of micro-changes in very specific contexts. What appears to be a metaphor may be the outcome of a number of changes in which pragmatic inferencing is activated in the flow of speech, as attested by textual data in historical corpora. The Invited Inferencing Theory of Semantic Change (IITSC) is a usage-based approach founded in the investigation of textual evidence (Traugott and Ko¨nig 1991; Traugott and Dasher 2002). Although sharing many of the assumptions of Cognitive Linguistics, it diverges in several respects, most notably in distinguishing between pragmatics and semantics. The term “invited inferencing” was borrowed from Geis and Zwicky (1971), but used in an extended way to evoke negotiation of meaning between speakers producing or even intending meanings beyond what is said, and hearers inferring such meaning. Likewise it draws on neo-Gricean pragmatics (Grice 1989; Horn 1984; Levinson 2000), and the distinction between particularized and generalized conversational implicatures, but appeals to partially different maxims. The Gricean Maxim that is considered key to change is his Quantity 2 (“Do not make your contribution more informative than is required”), combined with Relevance and rephrased as “Say no more than you must, and mean more thereby” (Horn 1984). Use of this maxim leads to rich interpretations. The hypothesis is that invited inferences that arise on the fly may become conventionalized (commonly activated) as generalized invited inferences. They may be “salient” in the community in that they can be drawn on consciously, cf. the causal implicature of after, but for the most part they are used unconsciously (Keller 1994). These generalized invited inferences may continue to be available over centuries, even millennia (cf. after). A regularly occurring context which “supports an inference-driven contextual enrichment” of one meaning to another has been called a “bridging context” (Evans and Wilkins 2000: 55). Conventionalization as a bridging context is a pragmatic development. Sometimes such inferences may be absorbed into the meaning of an expression with which they were formerly only pragmatically associated. In this case semantic reanalysis has occurred (Eckardt 2006) and a new coded meaning has become available, as evidenced by the use of an old form with the new meaning in a context which was not available before; e.g., siþþan ‘since’ was originally restricted to temporal ‘after’ and later came to be used with a causal meaning as well. In other words, semanticization of a formerly pragmatic meaning has occurred. Although originally discussed mainly with reference to grammaticalization, invited inferencing is conceived as a major motivation for semantic change in general (Traugott and Dasher 2002). It encompasses the changes associated with metonymy and metaphor, but also pejoration and amelioration. In the latter cases, the invited inferences are not only conceptually but also socially motivated. For example, those considered blessed and innocent may be evaluated as ignorant or foolish, cf. OE selig ‘blessed’ > ‘silly’, or Lt. nescius ‘unknowing, innocent’ (< ne + scius ‘not knowing’) > Middle Fr. nice ‘foolish’. Semanticization of a speaker’s beliefs and/or attitudes to what is being said is called “subjectification” (Traugott 1989, 2010; see also Davidse et al. [eds.] 2010). A partially overlapping, but more restricted, view of subjectification associates it with changes in the cognitive construal of vantage-point (e.g. Langacker 1990, 2006; see also Athanasiadou et al. [eds.] 2006). Subjectification in the first sense encompasses shifts not only from the perspective of the sujet d’e´nonce´ ‘syntactic subject’ to the sujet d’e´nonciation ‘speaking subject’ (Benveniste 1971), but also a range of meaning developments based
11 Linguistic Levels: Semantics and lexicon in the speaker’s perspective: spatial, temporal, metalinguistic, etc. Subjectification is extensively evidenced by the use of phrases like after all, anyway as discourse markers, and of adjectives like very (< ‘true’, see Fr. vrai) as scalar degree modifiers. It is also evidenced by the development of epistemic meanings, and of performative uses of speech act verbs. Many of the latter derive ultimately from past participles of Latin verbs, such as promise (< Lt. pro + miss- ‘forward sent’), suggest (< Lt. sub + gest‘under carried’) (note the conduit metaphors in the Latin, and the past participle form which shows these were originally stative and non-agentive).
4 Collocation and collostructional analysis Cognitive Linguistics and the Invited Inferencing Theory of Semantic Change are both conceptualized as “usage-based” theories. For the most part, the first is exemplified by constructed data, the latter by empirically attested data found in historical texts. Bybee has said of cognitive, usage-based linguistics in general that it is a “framework that allows change to be gradual and specific on various dimensions, such as the lexical, phonetic, and morphosyntactic, while at the same time providing general principles of linguistic organization that explain why change moves in certain directions and not others” (Bybee 2007: 981). Bybee draws attention especially to frequency effects on these dimensions. Particularly valuable in work on semantic change is the notion of collocation, or relationships among words or groups of words that go together. In a contextualized approach to the change in the meaning of conceit, for example, we find that from the beginning it was often associated with negative meanings. For example, of the five examples of conceit in Chaucer’s work, two are modified by wrong, and another is embedded in a context that suggests the opinion held is wanting: (1)
O sely preest! O sely innocent! With coveitise anon thou shalt be blent! O graceless, ful blynd is thy conceite (Canon’s Yeoman’s Tale 1076–1078) ‘Oh foolish priest! Oh foolish innocent! With covetousness you shall be blinded! Oh lacking God’s grace, fully blind is your opinion/mind’
Over time, although conceit could also be used in the sense of ‘good judgment’, the negative meaning became semanticized into conceit, presumably due to frequency of use in negative contexts. One of the interesting research questions is how some words become associated with either negative or positive contexts and whether or not these contexts become semanticized into the word. The phenomenon has been called “semantic prosody” (Stubbs 2001) and takes several shapes. One is illustrated by conceit, where a word is used so frequently in a negative context that the negative evaluation becomes part of the meaning. Others are preferred collocations, e.g. a shred of in the abstract quantifier sense ‘some’ has come to be used almost exclusively with not (hence it is understood as ‘not any’, and is often analyzed as a negative polarity item). Speakers often have preconceptions about such collocations, which may or may not be accurate. Invaluable for testing such preconceptions are computer-assisted
169
170
II Linguistic Levels approaches to corpora that provide “collostructional analysis”. Originally developed for synchronic work (Stefanowitsch and Gries 2003), this kind of analysis has been adapted for historical work on grammaticalization (Hilpert 2008), but could be extended to work on semantic change in general. It involves the exhaustive extraction of all tokens of a construction from a corpus while keeping one element constant, e.g. in the construction be going to+V, be going to with its verbal complements is kept constant. The objective is to determine not only which verbal complements it came to collocate with (this is “host-class expansion”, which Himmelmann 2004 considers criterial for grammaticalization), but also the strength of the association between them. It reveals changing selectional restrictions, hence fine-grained meaning change. It can be used to compare similar changes across languages, and to test schematic hypotheses about paths of change. For example, Hilpert (2008) tests Bybee et al.’s (1994: 270) hypotheses about the paths by which futures develop, one of them being (2): (2)
motion → intention → future
Collostructional evidence shows that Swedish komma at ‘come to’, although a motion verb that becomes a future marker, does not do so via intention. It also shows that in English be going to was initially used most often with speech act verbs like answer and begin, and became increasingly attracted to verbs with punctual meaning, especially those that are transitive with agentive meaning (e.g. get, marry), but also intransitive (e.g. die, leave). On the other hand, Dutch gaan ‘go’, which also became a future marker, was initially used primarily with motion verbs like laupen ‘walk’, and became increasingly attracted to intransitive, durative meanings (e.g. denken ‘think’, and voelen ‘feel’). Although both be going to and gaan exemplify (2), they have distinctly different semantic micro-histories, and hence different meanings.
5 Productivity of semantic changes at specific periods Not many studies have been conducted on semantic changes that affect a large class of items at a specific period of time. However, a few may be mentioned here. In one of the earliest attempts to demonstrate that semantic change can be patterned, Stern (1964 [1931]) showed that around 1300 adverbs meaning RAPIDLY developed the polysemy IMMEDIATELY in the context of perfective, i.e. punctual, verbs “denoting the action as a unit” (185–191), e.g. ME georne ‘rapidly’ > ‘immediately’. This change ceased, he said, around 1400, e.g. fleetly (1598) and rapidly (1727) itself did not undergo this change. Stern’s study is a non-quantified precursor of collostructional analysis, showing how change of meaning occurs metonymically in context. The change itself (which he calls a kind of “permutation”) is conceptualized as a case of what we would now call invited inferencing and unidirectionality of change: “it is evident that if a person rides rapidly up to another, the action is soon completed; but we cannot reverse the argument and say that if a person soon rides up to another, then the action is also rapidly performed” (Stern 1964: 186). A collostructional analysis would be needed to verify that this change was indeed as particular to the time as Stern claims. It is certainly crosslinguistically attested; Buck (1949: 964) commented that “the majority of words for ‘soon’ are, or were once, simply ‘quickly’ ” (for examples in Japanese see Traugott and Dasher 2002: 69). But that does not mean that there was a cluster of changes
11 Linguistic Levels: Semantics and lexicon that was highly productive at a particular point in time involving this particular semantic shift in any other language. Another example of a semantic change among adverbs that has been noted as having occurred particularly frequently at a given time is the development in Early Modern English of “boosters” or degree adverbs out of qualitative and highly evaluative adverbs (e.g. terribly, horribly, villainous – as now, some adverbs occurred without -ly). Peters (1994: 271) argues that most earlier boosters had developed from “local, dimensional or quantitative adverbs”, such as highly (dimensional) and vastly (quantitative). He hypothesizes that the development of boosters out of qualitative adverbs was associated with the development of more colloquial styles at the period. A third adverbial domain which is said to have become highly productive at a particular period is that of epistemic adverbs like probably, alledgedly, reportedly from the 18th century on. In recent work exploring the relationship between “cultural scripts” and semantics, Wierzbicka (2006) proposed that an increase in the number of epistemic adverbs is correlated with the advent of empiricism, especially under the influence of the philosopher John Locke’s An Essay Concerning Human Understanding (1690). Likewise, words like fairness and reasonableness underwent a significant shift. Using collocations as her evidence, Wierzbicka hypothesizes that our modern understanding of reasonableness includes common sense, a standard by which anyone’s behavior can be judged (108), and compatibility with reason (134), cf. reasonable doubt, reasonable care. However, prior to the enlightenment it meant ‘required by reason’ (134), or ‘having reason’, as in (3): (3)
Man is a resonable two foted beest (c.1380–87 Chaucer, Boece V. pr. iv. 128; Wierzbicka 2006: 109)
Testing Wierzbicka’s hypothesis that prior to the British enlightenment there was a cultural script of faith and certainty, Bromhead (2009) investigates a number of expressions like I think, in truth, verily, some of which, like verily, are no longer used, or recessive (in truth). She supports Wierzbicka’s conclusions in general, by providing evidence that the meanings of these expressions prior to the 18th century predominantly expressed certitude and confidence, rather than the doubt associated with the modern empiricist ethos. However, she also shows that the meaning changes Wierzbicka identifies were developing prior to the 18th century and the appearance of Locke’s book. This suggests that he was a synthesizer of current views, as well as a catalyst for their spread.
6 Differences between lexical and grammatical changes While there are similarities with respect to metaphor and metonymy between changes affecting lexical and grammatical (or grammaticalizing) expressions, there are also differences. Most notably, lexical semantic change concerns contentful change, whereas the development of grammatical expressions (see Section 2) is associated with “bleaching” or loss of contentful, lexical meaning: “[a]s grammatical morphemes develop, they lose specific features of meaning and thus are applicable in a wider range of environments” (Bybee 2007: 975). As Peters’s (1994) examples of boosters deriving from qualititative adverbs illustrates, grammaticalization may result in such collocations as
171
172
II Linguistic Levels terribly happy, or more recently pretty ugly. Bleaching is not found exclusively in grammaticalization. Occasionally lexical items may also lose in substantive content, e.g. OE þing ‘law court, assembly’ by metonymy > ‘thing, matter of concern’. But grammaticalization does not only involve loss of lexical meaning. There is enrichment of grammatical meaning as the original abstract implicature is semanticized, thus terribly lost the lexical meaning of ‘terror’, but gained abstract scalar meaning placing its complement high on the scale; pretty lost the lexical meaning ‘good-looking’, but gained scalar meaning, serving to place its complement above the median on a scale of intensity. Other respects in which semantic change associated with grammaticalization differs from that associated with lexical items is that it is more frequently replicated, and usually cross-linguistically attested. Subjectification occurs in lexical change (see Section 3 above) but it is particularly closely associated with grammaticalization because grammatical markers serve to indicate the speaker’s perspective on who does what to whom (case), how the situation is related to speech time (deictic tense) or to the temporality of a reference point other than speech time (relative tense), whether the situation is perspectivized as continuing and open-ended or not (aspect), whether the situation is relativized to the speaker’s beliefs (modality, mood), and how utterances are connected to each other (connectives, discourse markers), among other things (Traugott 2010).
7 Changes in the lexicon As indicated at the end of Section 1, the lexicon is the inventory of form-meaning pairs in a language. These pairs may be conceptualized as on a continuum from substantive, contentful, or “lexical”, like car, shoot, to “grammatical”, procedural, and indexical, like but, although. In the lexicon, meaning is often accounted for in terms of lexical relations, such as synonymy, antonymy, hyponymy/hyperonymy (taxonomic relations of member/superordinate set, e.g. carrots/vegetables), and meronymy/holonymy (part of/ whole of, e.g. finger/hand). Since the inventory of lexical and grammatical items that constitutes the lexicon consists of abstract structures without any inflections, the term “word” is not usually used in discussion of the lexicon. Instead, “lemma” is used for abstract lexemes, and sometimes “gram” for abstract grammatical items. It is usually assumed that the lexicon contains only items that are conventionalized in the sense that they are used by more than one speaker. However, it is particularly difficult in historical work on the earlier periods to know when this criterion is met, since if an item occurs only once, we cannot know whether this is because it was a nonce-item used by just one speaker, or whether it happens to appear in only one surviving manuscript. The debate has been particularly lively with respect to kennings in Old English. Some, especially those related to the sea, recur and may be considered candidates for the lexicon of Old English, e.g. swa¯nra¯d ‘swan road’ (see Kay, Chapter 20); many others may be nonce-forms. Kennings are usually thought of as metaphors constructed as compounds, but Broz (forthc.) analyzes them as complex metonymy-metaphor combinations. He illustrates with a productive process of forming kennings that denote the concept of ‘body’. In these, ba¯n ‘bone’ is the first element, as in ba¯ncofa ‘bone-chamber’, ba¯nfæt ‘bone-container’ or ba¯nhu¯s ‘bonehouse’. Broz suggests that several cognitive operations are at work here: a containment image schema which gives rise to conceptual metaphors such as BODY IS
11 Linguistic Levels: Semantics and lexicon CONTAINER (Lakoff 1987), which combines with the PART FOR WHOLE metonymy, bone being the essential part of the body. Changes affecting items in the lexicon clearly include semantic change, but research into the lexicon also concerns changes to the form of lexemes (e.g. hla¯f weard ‘loaf guardian’ > lord) (see Brinton, Volume 2, Chapter 100), and such factors as changes in the size of the lexicon and the consequences of borrowing. The latter two topics provide insight into what meanings were salient at particular periods, and in particular text-types, and will be the focus of this section. The size of the English lexicon at any period is hard to measure. The only criterion is the number of lemmas and grams listed in dictionaries, a factor highly dependent on tradition, and the purposes of a dictionary (see Kay, Chapter 20 and Dossena, Chapter 55 on some of the problems attendant on estimating increases in vocabulary size). Practices differ with respect to whether and when polysemies are counted as separate entries. When there is an academy that regulates language practice, as in France, a dictionary is less likely to include new words and meanings than when there is no such academy. That said, English speakers have been more willing than speakers of many other languages to borrow large amounts of vocabulary. The influence of Scandinavian, French, Latin, and Greek is widely known, but Arabic, Spanish, and more recently Japanese are among the many others that have also contributed to the inventory. No one speaker and no one register makes use of more than a tiny subset of all the possible lexical lemmas recorded in a dictionary. Dictionaries represent the aggregate of items used or usable by the totality of those regarded as speakers of a language. Since English speakers have been borrowing and inventing words for over a millennium and a half, and surprisingly few words are dropped from collective, as opposed to individual, inventories, the size of the vocabulary has been increasing since the beginning. As might be expected, such increases are not independent of periods of contact or new discourses and lifestyle changes (see borrowing in the 19th century of jihad for political and religious discourse, and of sushi for life-style discourse). Perhaps the most interesting increase, as shown by the Chronological English Dictionary, occurred in the period 1570–1630 (see discussion in Nevalainen 1999 and in Lancashire, Chapter 40). In this case it resulted not from contact, but from a conscious effort to shape English as a national language, no longer second to French or Latin, and from the explosion of new literary works, including the plays of Shakespeare. Many of the words he used never became conventionally used in English e.g. offendress ‘woman who offends’. Words were borrowed, especially if they were technical (cerebellum, specimen); derivational affixes were added to native or borrowed words (uncertitude; see in contemporary English u¨berlame); converted from one part of speech to another, e.g. from noun to verb (calendar ‘enter on one’s calendar’), or coined (giggle). When borrowing occurs, various kinds of consequences can be noted. Narrowing of one of the terms may occur if a native term already exists: compare the generic term cow (OE cu) with beef ‘kind of meat’ (Fr. boeuf ). Another is that the vocabulary may become multidimensionally stratified. Dimensions of variation that have received particular attention are region, social group, field of discourse (transaction, homily, letter), medium (spoken or written), and attitude (Kastovsky 2006). An example of social and age stratification is some contemporary teenagers’ use of intensifiers like dead, often in collocations that differ from those of standard varieties of English such as dead healthy, pure funny, enough funny (e.g. Stenstro¨m 2000; Macaulay 2005).
173
174
II Linguistic Levels Yet another consequence of borrowing is the coexistence of typologically different layers of vocabulary. Because many words have been borrowed into English from French, the lexicon abounds in words that have phonological alternations in the base (e.g. hı´story, histo´ric, historı´city), whereas native words do not (Kastovsky 2006: 212). The lexicon also includes verbs that have been said to be typologically very different from a semantic point of view. Different languages package semantic material into words in different ways (Lehrer 1992: 249). Talmy (e.g. 2000) proposed that there are two basic ways of expressing semantic path and manner/cause of motion in the verb. He distinguishes Chinese and all branches of Indo-European except Romance as languages that encode Motion and Manner/Cause together and treat Path as satellite. In Jill floated into the cave, float conflates Motion and Manner, in The napkin blew off the table, blow conflates Motion and Cause (wind). By contrast Romance, Semitic, Polynesian, and Navajo encode Motion and Path together and treat Manner as satellite, as in Spanish La botella entro´ a la cueva flotando ‘The bottle entered into the cave floating’. Because French words have been borrowed, we can use both enter (French), which treats Manner as satellite, and swim, which treats Path as satellite. Very interestingly, Latin, like English, encoded Motion and Manner/ Cause. Stolova (2008) attributes the change from Latin to Romance largely to the loss in Late Latin of verb prefixes that encoded Path, cf. abire ‘out-go’, and replacement of them in the Romance languages by verbs derived from nouns that had connotations of direction in them, e.g. Fr. monter ‘go up, climb’ < mons ‘mountain’. Verbs like entrare (< intra ‘into’) that had already become univerbated in Late Latin were not understood as encoding path morphologically, and so were treated like monter, as inherently including direction within them. This means that borrowings into English from Latin such as exit ‘outgo’, ascend (ad-scandere ‘at climb’) are typologically like English go up, while mount, enter from French are typologically different – we do not usually mount up a hill, though debts may mount up (a cumulative, not directional Path use of up), nor do we usually enter into a cave, although we may enter into an agreement, a conduit metaphor).
8 Future prospects Theoretical work on semantics, pragmatics, and the lexicon is increasing exponentially from many perspectives, among them computational, quantificational, typological, socio-cultural, and rhetorical. We can therefore expect substantial advances in research on semantic and lexical change in English in the near future. As these developments occur, the question of methodology in the use not only of electronic corpora, but also of dictionaries such as the OED will come to be of ever-increasing importance (see Allan and Robynson forthc.).
9 References Allan, Kathryn L. 2008. Metaphor and Metonymy: A Diachronic Approach. Malden, MA: WileyBlackwell. Allan, Kathryn and Justyna Robinson (eds.). forthc. Current Methods in Historical Semantics. Berlin/New York: De Gruyter Mouton. Anttila, Raimo. 1989 [1972]. Historical and Comparative Linguistics. Amsterdam/Philadelphia: Benjamins. Athanasiadou, Angeliki, Costas Canakis, and Bert Cornillie (eds.). 2006. Subjectification: Various Paths to Subjectivity. Berlin/New York: Mouton de Gruyter.
11 Linguistic Levels: Semantics and lexicon Barcelona, Antonio. 2000. On the plausibility of claiming a metonymic motivation for conceptual metaphor. In: Antonio Barcelona (ed.), Metaphor and Metonymy at the Crossroads: A Cognitive Perspective, 31–58. Berlin/New York: Mouton de Gruyter. Benveniste, Emile. 1971 [1958]. Subjectivity in language. In: Problems in General Linguistics, 223– 230. Trans. by Mary Elizabeth Meek. Coral Gables, FL: University of Miami Press. (Originally published as ‘De la subjectivite´ dans le langage’. Journal de psychologie 55 [1958]: 267–276.) Bre´al, Michel. 1897. Essai de Se´mantique (Science des Significations). Paris: Hachette. (English trans. by Mrs. Henry Cust, Semantics: Studies in the Science of Meaning. New York: Dover, 1900). Breban, Tine. 2008. Grammaticalization, subjectification and leftward movement of English adjectives of difference in the noun phrase. Folia Linguistica 42: 259–306. Bromhead, Helen. 2008. The Reign of Truth and Faith: Epistemic Expressions in 16th and 17th Century English. Berlin/New York: Mouton de Gruyter. Broz, Vlatko. forthc. Kennings: Riddles of metonymy or metaphor? In: Allan and Robynson (eds.). Buck, Carl Darling. 1949. A Dictionary of Selected Synonyms in the Principal Indo-European Languages. Chicago: University of Chicago Press. Bybee, Joan. 2007. Diachronic linguistics. In: Geeraerts and Cuyckens (eds.), 945–987. Bybee, Joan, Revere Perkins, and William Pagliuca. 1994. The Evolution of Grammar: Tense, Aspect, and Modality in the Languages of the World. Chicago: University of Chicago Press. Campbell, Lyle. 2004. Historical Linguistics. Cambridge, MA: MIT Press. Davidse, Kristin, Lieven Vandelanotte, and Hubert Cuyckens (eds.). 2010. Subjectification, Intersubjectification and Grammaticalization. Berlin/New York: De Gruyter Mouton. Dı´az Vera, Javier E. (ed.). 2002. A Changing World of Words: Studies in English Historical Lexicography, Lexicology and Semantics. Amsterdam/New York: Rodopi. Eckardt, Regine. 2006. Meaning Change in Grammaticalization: An Enquiry into Semantic Reanalysis. Oxford: Oxford University Press. Evans, Nicholas and David Wilkins. 2000. In the mind’s ear: The semantic extensions of perception verbs in Australian languages. Language 76: 546–592. Fortson, Benjamin W., IV. 2003. An approach to semantic change. In: Brian D. Joseph and Richard D. Janda (eds.), The Handbook of Historical Linguistics, 648–666. Oxford: Blackwell. Geeraerts, Dirk. 1997. Diachronic Prototype Semantics: A Contribution to Historical Lexicology. Oxford: Clarendon Press. Geeraerts, Dirk and Hubert Cuyckens. 2007. Introducing Cognitive Linguistics. In: Geeraerts and Cuyckens (eds.), 3–21. Geeraerts, Dirk and Hubert Cuyckens (eds.). 2007. The Oxford Handbook of Cognitive Linguistics. Oxford: Oxford University Press. Geis, Michael L. and Arnold M. Zwicky. 1971. On invited inferences. Linguistic Inquiry 2: 561–566. Grice, H. Paul. 1989. Logic and conversation. Studies in the Way of Words, 22–40. Cambridge, MA/ London: Harvard University Press. (first published in Peter Cole and Jerry L. Morgan [eds.], Syntax and Semantics III: Speech Acts, 41–58. New York: Academic Press, 1975). Grondelaers, Stefan, Dirk Speelman, and Dirk Geeraerts. 2007. Lexical variation and change. In: Geeraerts and Cuyckens (eds.), 988–1011. Haspelmath, Martin. 1997. Temporal Adverbials in the World’s Languages. Munich: LINCOM EUROPA. Heine, Bernd, Ulrike Claudi, and Friederike Hu¨nnemeyer. 1991. Grammaticalization: A Conceptual Framework. Chicago: University of Chicago Press. Heine, Bernd and Tania Kuteva. 2002. World Lexicon of Grammaticalization. Cambridge: Cambridge University Press. Hilpert, Martin. 2008. Germanic Future Constructions: A Usage-based Approach to Language Change. Amsterdam/ Philadelphia: John Benjamins. Himmelmann, Nikolaus P. 2004. Lexicalization and grammaticalization: Opposite or orthogonal? In: Walter Bisang, Nikolaus P. Himmelmann, and Bjo¨rn Wiemer (eds.), What Makes Grammaticalization – A Look from its Fringes and its Components, 19–40. Berlin/New York: Mouton de Gruyter.
175
176
II Linguistic Levels Hock, Hans Henrich and Brian D. Joseph. 1996. Language History, Language Change, and Language Relationship: An Introduction to Historical and Comparative Linguistics. Berlin/New York: Mouton de Gruyter. Horn, Laurence R. 1984. Toward a new taxonomy for pragmatic inference: Q-based and R-based implicature. In: Deborah Schiffrin (ed.), Meaning, Form, and Use in Context: Linguistic Applications, 11–42. (Georgetown University Round Table ’84.) Washington, DC: Georgetown University Press. Kastovsky, Dieter. 2006. Vocabulary. In: Richard Hogg and David Denison (eds.), A History of the English Language, 199–270. Cambridge: Cambridge University Press. Kay, Christian, Jane Roberts, Michael Samuels, and Irine´ Wotherspoon (eds.). 2009. Historical Thesaurus of the Oxford English Dictionary. Oxford: Oxford University Press. Keller, Rudi. 1994. On Language Change: The Invisible Hand in Language. London: Routledge. (Originally published as Sprachwandel: Von der unsichtbaren Hand in der Sprache. Tu¨bingen: Francke, 1990.) Ko¨vecses, Zolta´n and Gu¨nter Radden. 1998. Metonymy: Developing a cognitive linguistic view. Cognitive Linguistics 9: 37–77. Krug, Manfred G. 2000. Emerging English Modals: A Corpus-based Approach to Grammaticalization. Berlin/New York: Mouton de Gruyter. Lakoff, George. 1987. Women, Fire, and Dangerous Things: What Categories Reveal about the Mind. Chicago: University of Chicago Press. Langacker, Ronald W. 1990. Subjectification. Cognitive Linguistics 1: 5–38. Langacker, Ronald W. 2006. Subjectification, grammaticalization, and conceptual archetypes. In: Athanasiadou, Canakis, and Cornillie (eds.), 17–40. Lehrer, Adrienne. 1992. A theory of vocabulary structure: Retrospectives and prospectives. In: Mario Pu¨tz (ed.), Thirty Years of Linguistic Evolution: Studies in Honour of Rene´ Dirven on the Occasion of his Sixtieth Birthday, 243–256. Berlin/New York: Mouton de Gruyter. Levinson, Stephen C. 2000. Presumptive Meanings: The Theory of Generalized Conversational Implicature. Cambridge, MA: MIT Press. Macaulay, Ronald K. S. 2005. Talk that Counts: Age, Gender, and Social Class Differences in Discourse. Oxford: Oxford University Press. Nevalainen, Terttu. 1999. Lexis and semantics. In: Roger Lass (ed.), The Cambridge History of the English Language. Vol. III. 1476–1776, 332–458. Cambridge: Cambridge University Press. Panther, Klaus-Uwe and Linda L. Thornburg. 2003. Metonymies as natural inference and activation schemas: The case of dependent clauses as independent speech acts. In: Klaus-Uwe Panther and Linda L. Thornburg (eds.), Metonymy and Pragmatic Inferencing, 127–147. Amsterdam/ Philadelphia: Benjamins. Peters, Hans. 1994. Degree adverbs in Early Modern English. In: Dieter Kastovsky (ed.), Studies in Early Modern English, 269–288. Berlin/New York: Mouton de Gruyter. Pustejovsky, J. 1995. The Generative Lexicon. Cambridge, MA: MIT Press. Reddy, Michael J. 1993 [1979]. The conduit metaphor: A case of frame conflict in our language about language. In: Andrew Ortony (ed.), Metaphor and Thought, 164–201. Cambridge: Cambridge University Press. Sperber, Dan and Dierdre Wilson. 1995. Relevance: Communication and Cognition. 2nd edn. Oxford: Blackwell. Stefanowitsch, Anatol and Stefan Th. Gries. 2003. Collostructions: Investigating the interaction between words and constructions. International Journal of Corpus Linguistics 8: 209–243. Stenstro¨m, Anna-Britta. 2000. It’s enough funny, man: Intensifiers in teenage talk. In: John M. Kirk (ed.), Corpora Galore: Analyses and Techniques in Describing English, 177–190. Amsterdam/Atlanta, GA: Rodopi. Stern, Gustaf. 1964. Meaning and Change of Meaning: With Special Reference to the English Language. Bloomington/London: Indiana University Press. (Reprint of Elanders Boktryckeri Aktiebolag, Go¨teborg, 1931)
12 Linguistic Levels: Idioms and fixed expressions
177
Stolova, Natalya I. 2008. From satellite-framed Latin to verb-framed Romance: Late Latin as an intermediate stage. In: Roger Wright (ed.), Latin vulgaire, latin tardif: Actes du VIIIe`me Colloque International sur le Latin Vulgaire et Tardif, Oxford, 6–7 Septembre 2006, 253–262. Hildesheim: Georg Olms Verlag. Stubbs, Michael. 2001. Words and Phrases: Corpus Studies of Lexical Semantics. Oxford: Blackwell. Sweetser, Eve E. 1990. From Etymology to Pragmatics: Metaphorical and Cultural Aspects of Semantic Structure. Cambridge: Cambridge University Press. Talmy, Leonard. 1988. Force dynamics in language and cognition. Cognitive Science 2: 49–100. Talmy, Leonard. 2000. Toward a Cognitive Semantics. 2 vols. Cambridge: Cambridge University Press. Traugott, Elizabeth Closs. 1989. On the rise of epistemic meanings in English: An example of subjectification in semantic change. Language 65: 31–55. Traugott, Elizabeth Closs. 2010. (Inter)subjectivity and (inter)subjectification: A reassessment. In: Davidse, Vandelanotte, and Cuyckens (eds.), 29–70. Traugott, Elizabeth Closs and Richard B. Dasher. 2002. Regularity in Semantic Change. Cambridge: Cambridge University Press. Traugott, Elizabeth Closs and Ekkehard Ko¨nig. 1991. The semantics-pragmatics of grammaticalization revisited. In: Elizabeth Closs Traugott and Bernd Heine (eds.), Grammaticalization, Vol. 1, 189–218. Amsterdam/Philadelphia: John Benjamins. Ullmann, Stephen. 1964. Semantics; An Introduction to the Science of Meaning. Oxford: Blackwell. Warren, Beatrice. 1992. Sense-Developments: A Contrastive Study of the Development of Slang Sense and Novel Standard Sense in English. Stockholm: Almqvist & Wiksell. Warren, Beatrice. 1999. Laws of thought, knowledge and lexical change. In: Andreas Blank and Peter Koch (eds.), Historical Semantics and Cognition, 215–243. Berlin/New York: Mouton de Gruyter. Wierzbicka, Anna. 2006. English: Meaning and Culture. Oxford: Oxford University Press.
Elizabeth Closs Traugott, Palo Alto (USA)
12 Linguistic Levels: Idioms and fixed expressions 1 2 3 4 5 6 7 8
Introduction The concept and scope of phraseology and phraseological research Approaching historical English phraseology Metalinguistic sources and their value for the identification of phraseological units in historical texts Origin, change, and earlier uses of English phraseological units The impact of phraseology and collocation / string frequency on language change Summary References
Abstract In order to establish a framework of reference for an approach to historical English phraseology, a brief discussion of the concept and scope of phraseology and phraseological scholarship is provided in Section 2. On this basis, Section 3 approaches historical phraseology in two consecutive steps – always paying special regard to English: an overview of the state of historical phraseology leads to major approaches, which establish the main Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 177–196
178
II Linguistic Levels procedure in Sections 4 to 6. In Section 4, the role of metalinguistic sources for historical phraseology is pointed out, especially with relation to the identification of phraseological units in historical texts. Theories of the origin, rise, and kinds of change of phraseological units as well as the employment of English phraseological units in literary texts are the subjects of Section 5. While Section 5 is thus concerned with change in the phraseological system over time, Section 6, vice versa, finishes off the article with an outlook on the influence of frequent word combinations and phraseology on language change in a broader perspective.
1 Introduction Since about the middle of the 20th century, a branch of linguistics that is concerned with the study of “idioms and fixed expressions” has developed in Europe under the name of “phraseology”. Being the oldest and most comprehensive systematic branch of linguistics dealing with “idioms and fixed expressions”, or “formulaic language” – which, however, still lacks a fully developed historical approach – it will provide the point of departure for this article. Owing to the developments in American and British linguistics in the past century, the branch of phraseology has only recently started to gain a footing there and to complement existing studies on specific aspects of formulaic language, such as idioms and conversational routines. In fact, the year 2007 stands out with a number of pertinent publications. However, only very few historical studies of English phraseology have been published as yet, while some publications which were not explicitly conceived as belonging to this branch can certainly be viewed from this perspective.
2 The concept and scope of phraseology and phraseological research The concept and scope of phraseology will be discussed by first approaching phraseology through formulating a description of phraseological units as its objects of study, giving some examples, and pointing out defining criteria and major properties of these units, in particular highlighting the factor of variation. In the main, this subchapter proceeds along lines which now seem to be accepted among phraseologists (cf., e.g., Cowie 1998) after a long period of heated discussion (cf., e.g., Welte 1992). Having thus outlined its content, one major classification of English phraseology that has been suggested will briefly be referred to in order to be able to locate findings of individual (historical) studies within a descriptive system of phraseology. Finally, a rough outline of established approaches of phraseological research will be briefly indicated in order to be able to determine a place for historical (English) phraseology.
2.1 Defining criteria and major properties of phraseological units “Phraseological unit” seems to be one of the most widely accepted English terms for the linguistic units studied by scholars of phraseology. A description may be given as follows: Phraseological units are semantically and/or pragmatically fixed units in a language which consist of two or more smaller units that are either lexemes of that language
12 Linguistic Levels: Idioms and fixed expressions outside these expressions or recognized as unique lexemes within them, and that together do not exceed sentence length. Therefore, high frequency structural units such as the phraseological unit as well in example (1), and pragmatic markers as for instance in (2), belong to phraseology just like infrequent idioms as in (3), or proverbs as in (4). (1)
as well ‘too; also’ (All meaning equivalents quoted in this article to explain phraseological units are taken from Cowie et al. (1983) and Cowie and Mackin (1993), if not stated otherwise.)
(2)
you know ‘I am informing, or reminding, you’ (but cf. also Brinton [1996: 42], who points out the following functions of you know: ‘indicating knowledge shared between speaker and hearer’, ‘indicating general knowledge’, and ‘presenting new information as if it were old information in order to improve its reception’)
(3)
jobs for the boys ‘the provision of paid employment for favoured groups within a hierarchy, profession, administration etc. (the implication being that the work of these groups is not really necessary)’
(4)
Let sleeping dogs lie ‘do not provoke, disturb or interfere with somebody/something that is giving no trouble though he/it might, or could, do so’
In linguistic analysis, phraseological units may be marked with regard to their semantics as in the much-quoted opaque expression (5), their lexis (6), syntax (7), or pragmatics (8). (5)
to kick the bucket ‘(informal) die’ (not equivalent to ‘to move one’s foot violently against a pail’)
(6)
to put the kibosh on something/somebody ‘dispose of finally, finish off, do for’ (cf. Oxford English Dictionary (Simpson [ed.] 2000–; henceforth OED), s.v. kibosh n.; the word kibosh (of obscure origin) is not used in general English)
(7)
to trip/dance/tread the light fantastic ‘(facetious) dance’ (< Come, and trip it as ye go / On the light fantastic toe, from John Milton’s L’Allegro; the phraseological unit lacks the noun)
(8)
an eye for an eye (and a tooth for a tooth) ‘(a warning that) an act of aggression will be met with retaliation of the same kind (especially in personal or national conflicts)’ (Exodus xxi. 22–24; the meaning is situationally determined)
Phraseological units may carry connotations and may have intensifying and emphatic functions in a text, as Gla¨ser (1986, 1998) points out from the point of view of stylistic
179
180
II Linguistic Levels analysis, and they can be fruitfully studied from a cultural perspective, too (cf. the contributions in Skandera 2007; for a recent study of the special features of phraseological units, cf. Dobrovo 0skij and Piirainen 2009). For their linguistic study it is important to consider that phraseological units, once they are implemented in a speech community, are reproduced in language use rather than appearing as newly produced structures each time they are employed. What makes them difficult to describe and may render them very hard to identify by means of computer processing in, especially historical, texts are at least four major factors: Phraseological units are idiomatic to varying degrees; idiomaticity is here understood semantically, referring to the relation of the meanings of the parts to the meaning of the whole expression (semantic [non-]compositionality). Cowie (1983: xiii) distinguishes between three broad categories of phraseological units primarily on the basis of semantic compositionality. Firstly, there are fully idiomatic phraseological units (pure idioms) such as example (5); they have a fixed form and no literal meaning. Secondly, figurative idioms such as the first example in (10) possess a literal and a figurative meaning. Thirdly, restricted collocations such as example (9) are characterized by a combination of one word which appears in a literal sense and another whose meaning is determined by the context. While some members of the third category allow lexical variation, it is rare in the second. (9)
to jog one’s/somebody’s memory ‘remind somebody of / about something; help or stimulate somebody to recall something’ (to jog is used in a transferred sense, memory is used literally)
The frequency of use differs for various types of phraseological units (cf. especially Moon 1998b: 88–89). Phraseological units may show unpredictable transformational behavior, for instance with regard to passivization (10). However, the syntactic behavior of individual expressions follows (strong) tendencies rather than strict rules (cf., e.g., Moon 1998b: 88–90). Thus, while the expression (10a) can be transformed (“⇒”) into (10b), for example, there is a strong tendency that the phraseological unit (10c) cannot be transformed into the passive (10d). (10) a. to spill the beans ‘(informal) give away information, deliberately or unintentionally’ ⇒ b. the beans were spilled; BUT c. to drive a hard bargain ‘(have the means, power, or cunning to) force a bargain; contrive an exchange of goods and services that is either unfair or to one’s own advantage’ ⇒ d. *a hard bargain is driven (the passive transformation is not possible according to Cowie et al. 1983; however, an unmonitored Google search in December 2011 for the expression in the affirmative, including different grammatical forms of the verb phrase, yielded over 3 million hits for the active construction, but also some 560 hits for the passive one, among them the following quotation from The Telegraph of 8 November 2006, “‘A hard bargain was driven over appearances,’ said Jennings of the transfer which paid an initial £5 million.” http:// www.telegraph.co.uk/sport/columnists/henrywinter/2349920/Walcott-well-upto-speed-on-his-learning-curve.html)
12 Linguistic Levels: Idioms and fixed expressions Above all, phraseological units possess different degrees of formal fixedness (see especially Moon 1998a: 120–177 and Moon 1998b: 90–96; see also the paragraph on idiomaticity, above): in addition to the realization of, e.g., one’s/somebody’s in (9), which depends on the context, and lexical variation and choice in (7) and (11), for instance, phraseological units may also be open to modification and show a striking breadth of creative employment for particular purposes. This may happen to such an extent that even an almost wholly deviant set of words may still be related to one particular phraseological unit (12) (cf. Fiedler 2007: 95–96; Gla¨ser 1998: 142–143). Along similar lines, finally, phraseological patterns such as the combination of subject and predicative constituent in the “Incredulity Response Construction” as exemplified in (13) are constructions with highly idiosyncratic meaning but not much lexical fixation (cf. Fleischer 1997: 130; Sailer 2007: 1069; Huddleston and Pullum 2002: 890, who term these “bare predication polar echo constructions”). This lack of fixedness may make phraseological units difficult to handle by both the speaker and the linguist. In addition, scholars of phraseology have to deal with individual, diastratic, and diatopic variation in a language’s store of phraseological units (“phrasicon”). (11) cap/hat in hand ‘uncovering the head as a sign of reverence, respect, courtesy’ (cf. OED, s.v. cap n.1, def. 4h and hat n., def. 5a; cap in hand in use since the 16th century, hat in hand attested in the 19th century) (12) Others’ Trash Is Now An Architect’s Treasure (The New York Times / Su¨ddeutsche Zeitung 13 June 2005; example taken from Fiedler 2007: 96. The sentence’s semantics, formal structure, and rhythm unambiguously relate it to the proverb One man’s meat is another man’s poison ‘what seems good or pleasing to one person may be bad or unsuitable for another’.) (13) Me worry?! Kim resign?! Her a genius?! (response construction expressing incredulity)
2.2 Classifying phraseological units In the light of such a wide and complex range of criteria pertaining to the characterization of phraseological units, it follows that in order to ascertain the scope of phraseology, a comprehensive, “neutral” descriptive framework of reference is called for. Rosemarie Gla¨ser’s attempt of 1986 will be introduced at this point (for brief versions in English, cf. Gla¨ser 1998; Knappe 2004: 15–22). Following the earlier Russian tradition and the discussion based on it, Gla¨ser adapts the system to English and distinguishes (beyond the polar notion of idiomaticity and non-idiomaticity as the primary division) between “word-like” units, called “nominations”, which form the centre of English phraseology, and “sentence-like” units, called “propositions”, which are found on the periphery. The nominations are classified according to word class and include, for instance, nominal phraseological units such as (3) and verbal phraseological units as found in (5), (6), (7), (9), and (10). An example of a proposition is, for instance, the proverb (4), but slogans, commandments, maxims, routine formulae, and quotations belong in this category, too. Between the nominations and the propositions, Gla¨ser establishes a class of “reduced propositions” which are nominations in form but on a
181
182
II Linguistic Levels deeper level are propositions, such as the fragment of a proverb (14) and the stereotyped comparison (15). These can overlap with the nominations, as in the case of irreversible binomials with and, which in (16) should be viewed as one nominal phraseological unit, and sometimes a fragment of a proverb may be hard to identify for native speakers. Although this system is rather comprehensive, further categories can be suggested (cf., e.g., Burger 2010: 33–58). (14) a new broom (< a new broom sweeps clean) ‘somebody recently appointed to office or a responsible post’ (starts with an energetic programme of reform and change, sometimes not welcomed by those already there) (15) as blind as a bat [is] ‘unable to see, or read, very easily’ (but usually not completely blind); (figuratively) unable to see, or perceive, something that is obvious to other people (two underlying propositions in, e.g., He is as blind as a bat: He is blind and A bat is blind) (16) bread and butter ‘livelihood’ (Gla¨ser 1986: 46) (Sub-)classifications of phraseology may of course fruitfully be tailored to the scholars’ research objectives. In order to give just one example, which is also interesting for historical lexicography, Rosamund Moon’s (1998a, 1998b) approach will be briefly referred to at this point. She works from a lexicographer’s perspective and uses corpus linguistic methods to study “fixed expressions and idioms”. Her typology answers the question of which phraseological entries should be included in a monolingual dictionary and to this end addresses three different kinds of non-compositionality in phraseological units: problems of lexico-grammar are inherent in anomalous collocations, such as the “cranberry collocation” in (6), dependence on the discoursal or situational interpretation marks the class of formulae, which is exemplified, for instance, by the “saying” (8), and finally three degrees of semantic transparency are distinguished, (5) being an example of an “opaque metaphor”. General difficulties of classifying phraseological units beyond their structure or syntactic function are apparent in Moon’s finding that one fourth of the c.6,700 phraseological units which she considers fall into more than one of these categories.
2.3 The study of phraseology, with particular regard to English Within the framework of phraseological scholarship, phraseological units have been investigated from a wide array of perspectives, which cannot be aptly summarized in this article. A handy overview of established approaches may be found in the 20 chapters of the two-volume Phraseologie/Phraseology, published in the series Handbooks of Linguistics and Communication Science (Burger et al. 2007). One important observation is that the number of English contributions in handbooks and collections of papers has increased considerably in recent years. This will certainly further research in English phraseology, as all scholars of English are addressed by the use of this metalanguage (from this perspective, compare Burger et al. 2007 to Burger et al. 1982; cf. also the collections of English articles by Cowie 1998 and Skandera 2007).
12 Linguistic Levels: Idioms and fixed expressions Norrick (2007) has summarized the state of phraseological research within the main linguistic theories and approaches of modern Anglo-American linguistics (cf. also Sick 1993: 17–54): He found that systematic treatment of phraseological units is rare (Norrick only mentions Makkai 1972; but systematic classifications can also be found in some other studies); American research discusses the issue of idioms within a generative framework; in the British tradition, collocation studies and corpus linguistics have contributed to phraseological research. On the other hand, the studies of English phraseology by Cowie (1983), Gla¨ser (1986, 1998) and Moon (1998a), as well as several contributions in Cowie (1998) such as those by Altenberg, Mel 0cˇuk, and Howarth, and also the monograph Idiom Structure in English by Makkai (1972), are to varying degrees indebted to the (Eastern) tradition of phraseology. Further important publications include, for example, studies from the perspectives of (first and second) language acquisition (e.g. Wray 2002 from a theoretical standpoint, and Nattinger and DeCarrico 1992 from the point of view of language teaching), proverbs (paremiology), pragmatics, metaphor, and psycholinguistics as well as cognitive linguistics, the matter of irreversible binomials and factors influencing the fixedness of word groups (cf. Hudson 1998), to mention but a short selection. Some of these studies show that different theoretical approaches such as stratificational grammar and idiom studies (Makkai), theories of the mental lexicon and formulaic language (Wray), or practical tools such as corpus linguistics and the investigation of phraseology (Moon, Altenberg, Mel 0cˇuk) can be combined in trying to explore the nature and use of idioms and fixed expressions in English.
3 Approaching historical English phraseology While the phraseological units of present-day English have been given increasing attention, the exploration of historical phraseology has not increased in like degree. In fact, it is still correct to speak of a Cinderella status of historical (English) phraseology: As far as the English collections are concerned, they do not include a single historical article apart from one contribution on the development of English proverb collections (Doyle 2007b; this is the sole focus of Doyle 2007a, too). In book-length studies of English phraseology, concern with the origin and diachronic change of phraseological units is expressed only briefly if at all: thus Welte (1992: 578) judges the historical aspect to be “fascinating” but does not elaborate on it, and Gla¨ser (1986: 51–53) devotes three pages to “diachronic aspects” of phraseology which owe their inclusion in her study of Phraseologie der englischen Sprache exclusively to their relevance for present-day usage, with emphasis on relic forms, idiomatization, and word-formation. In Fiedler’s (2007) coursebook, historical phraseology has no place at all. Where historical phraseology is treated in its own right in comprehensive surveys and collections – as in Burger et al. (1982: 315–382), Fleischer (1997: 244–246), Ha¨cki Buhofer and Burger (2006: 413– 465) and also in the new collection by Burger et al. (2007: 1078–1145, with contributions on English, German, French, Italian and the Slavic languages) – it is striking that it tends to be the final chapter (not, however, in Burger 2010: 129–154 and Thun 1978: 75–84). The systematic inclusion of historical phraseology in overviews may, however, be interpreted as a sign that scholars are getting ready to explore the place of phraseology in the history of a language in its own right. Historical phraseology with its two branches
183
184
II Linguistic Levels of synchronic investigation of past systems and diachronic investigation into the development of the language data both re-opens the panorama of phraseological research and at the same time demands methods particularly tailored to historical study. As far as studies specifically devoted to historical English phraseology are concerned, published studies include recent papers such as Howarth’s (2000) contribution, which highlights pertinent aspects of phraseological variation and change around a model of conventionality, concentrating on the period 1800–1900, and Cowie (2003) tests the possibilities of the OED for the study of English (historical) phraseology. Dictionaries of idioms and proverbs have also contributed to the study of English historical phraseology (cf. Section 4.2, below). Monographs and collected volumes on aspects of historical English phraseology are rare: examples are Hiltunen (1983), Brinton and Akimoto (1999), Claridge (2000), Moralejo Ga´rate (2003), Trousdale (2008) within the framework of construction grammar – all on aspects of “multi-word verbs”, in particular on the rise of “complex” and phrasal verbs; Prins (1952) on French influence on English phrases; Sontheim (1972) on proverbs and proverbial expressions; Weinstock (1966) on the use of formulaic language in Shakespeare; Knappe (2004) on phraseology in English language study to 1800. A large-scale survey of historical English phraseology has not been published yet. In fact, a great number of research questions are waiting to be explored. These are too numerous to be listed explicitly and comprehensively in this article. Rather, it is hoped that the following systematic (but necessarily not comprehensive) account of possibilities of English historical phraseology, which are presented together with selected references to existing studies, will help to lay a basis for its future exploration within an accepted system of “idioms and fixed expressions” (phraseology). Three major approaches to historical phraseology have been singled out in the past (cf. in particular Sialm et al. 1982; Burger and Linke 1998; also Howarth 2000), which will be looked at in the following sections: a. the study of metalinguistic sources such as proverb collections, dictionaries, and grammars and their contribution to the identification of phraseological units in historical texts (Section 4); b. the investigation of the origin and change of phraseological units (Section 5.1); and c. their usage and function in earlier texts (Section 5.2). Another promising area for future research is the study of the impact of phraseological units and frequent collocations on larger issues of language change (Section 6).
4 Metalinguistic sources and their value for the identification of phraseological units in historical texts Historical metalinguistic sources are relevant for historical English phraseology in a twofold way: first, as representing the scholarship of past ages they are the objects of the historiography of English phraseology both in its own right and with regard to the judgment of the reliability of the works as data sources. Second, both historical and modern collections of historical phraseological language data are, if they are judged to reflect actual language use, major aids in the identification of phraseological units in historical texts.
12 Linguistic Levels: Idioms and fixed expressions
4.1 The historiography of English phraseology A book-length survey of the place of phraseology in pre-19th-century English language study is Knappe (2004). Included are: English proverbs in – – special collections (cf. now also Doyle 2007a, who goes beyond 1800); and – – textbooks; as well as English phraseological units in – – stylistic treatises; – – handbooks for foreign-language teaching and contrastive language study; – – bilingual and multilingual lexicography with English as the first language entered; – – translation theory; – – philosophical and universal language schemes (including shorthand writing); – – monolingual English lexicography; – – monolingual English grammars; and – – early classifications of phraseological units. The analysis of the English material showed that particular types of phraseological units stand out as being especially interesting to the early textbook authors. These are proverbs, routine formulae, stereotyped comparisons, binomials, restricted collocations, phrasal verbs, figurative idioms, and idiomatic phraseological units in general. As early as from the time of the Restoration, and partly owing to the endeavours of the Royal Society (especially members of the Society such as John Ray and John Wilkins), scholars writing in the British Isles started approaching the topic in a more analytical way, developing classifications of the phraseological material (cf. also Knappe 2006c) which culminated in George Campbell’s Philosophy of Rhetoric (1776). Joseph Priestley in the 18th century can be credited with the emancipation of the notion of an “idiom” from its roots in language comparison. For the identification of phraseological units in historical texts it is important to know that the material discussed or entered in these textbooks and dictionaries reflects actual language use (cf. also Section 4.2, below). It is hoped that future studies will extend and complement the findings in Knappe (2004) and join them up with the beginning of the scholarly study of phraseology in the 20th century. In the long run, the history of English phraseology will have to be viewed in a European perspective (for German, cf. also Burger et al. 1982: 360–382 and Weickert 1997).
4.2 The identification of phraseological units in historical texts Several criteria have been singled out that may lead to a proper judgment of whether a given historical word combination – found by targeted reading, corpus search, or a combination of the two – may claim phraseological status (cf., e.g., Friedrich 2007: 1093; Burger in Sialm et al. 1982: 346-360). These criteria may be deduced both from linguistic considerations and from metalinguistic sources. Thus, a linguistic criterion may be the fact that a modern parallel of the expression exists, as in (11), the communicative function may betray a routine formula (17), formal criteria such as alliteration are typical of some expressions (18), word-formations are sometimes based on earlier
185
186
II Linguistic Levels phraseological units, such as the dephraseological formation in (19), the degree of idiomaticity (semantic non-compositionality) of the whole expression is a strong indicator of phraseological status, as in (20), and there are cases where the original composition of the phraseological unit has become opaque (21). The exact meaning of the components at particular times can be difficult to determine, and thus it may be hard to distinguish idioms from restricted collocations in past stages of the language (22). Finally, in the case of translated texts, word combinations which are (recurrently) different from the original may be further investigated for their phraseological status, and the high frequency of occurrence of structural units or pragmatic markers is an indication of phraseological status, too, as it is in present-day language. (17) How do you? ‘How are you?’ “God be thanked for you, How do you?”, quotation 1570 (cf. OED, s.v. do v., def. 19) > How do you do? (cf. OED, s.v. how adv., def. 2a and how-do-you-do, how-d’ye-do phr. and n.; first quotation: 1697) (18) kith and kin ‘country and kinsfolk’ > ‘acquaintance and kinsfolk’ > ‘relatives’ (cf. OED, s.v. kith n., def. 5) (19) fence-sitter < to sit on the fence (cf. Gla¨ser 1986: 52; cf. OED, s.v. fence n., def. C2 and sit v., def. B3c, in quotation 1887: “Those who sit ‘on the fence’ – men with impartial minds, who wait to see … ‘how the cat will jump’ ”; first occurrences: fence-sitter 1905, slightly earlier: fencesitting 1904) (20) to bear in/on hand ‘bring forth something wrong against someone’ (14th–16th century; cf. OED, MED, early dictionaries) (21) good-bye < God be with you/ye (cf. OED, s.v. good-bye) (22) at non hond ‘in no way, not at all’, figurative use of Middle English hond as ‘a position or direction to one side or the other’? (cf. MED, s.v. ho¯nd(e n., def. 7; last quotation of at no hand in OED, s.v. hand n1., def. 25g: 1690) Metalinguistic indications, such as the comments “literally” or “proverbial” in a text (e.g. 1835: “We came in literally neck and neck”; 1880: “The Burials Bill is thought to resemble the proverbial chip in porridge, which does neither good nor harm”; Howarth 2000: 226), or the inverted commas in the quotation in (19), are helpful for the identification of phraseological units to a certain extent, but reliable larger collections, which we possess from the 17th century onwards (and for proverbs also earlier than this), are certainly profitable resources for historical English phraseology: it was not only the analytic spirit which developed in Restoration England, but this period may also be seen as the time when collectors started to restrict themselves to the inclusion of authentic
12 Linguistic Levels: Idioms and fixed expressions English language material. Thus, from the 17th century onwards contrastive handbooks for foreign language teaching and dictionaries stand out as rich and to a large extent unexploited storehouses of language data, which in the future will be highly useful in the investigation of the historical development of English phraseological units. Late 19th, 20th, and early 21st-century historical lexicography, in particular the OED and the Middle English Dictionary (Kurath et al. 1952–2001; henceforth MED), which both are now searchable online, as well as efforts of phraseography (dictionaries of idioms and fixed expressions) and especially paremiography (dictionaries of proverbs) (cf. Apperson 1929; Tilley 1950; Whiting 1968; Mieder et al. 1992; Speake 2003) provide information on language data which is of paramount importance for the study of the historical development of phraseological units. As far as the great OED is concerned, Cowie (2003) found that much phraseological material can be retrieved from its text but that this material is often neither prominently displayed nor consistently handled. Earlier alphabetic and topical dictionaries can usefully complement these modern lexicographical endeavours, as has for instance been shown in a pilot study on Roget’s Thesaurus (cf. esp. Knappe 2006a, but also 2006b) – although here the exact meaning and potential formal variations of a given phraseological unit may not be found in the collections, but will have to be deduced from actual texts. Reliable and comprehensive collections of phraseological units from all periods of English are most important for the study of historical phraseology. Such an historical database of phraseological units can then fruitfully be employed as the point of departure for corpus-based investigation along the lines of the approach adopted in Moon’s research (1998a, 1998b), who gathered her material on the basis of the phraseological entries in the first edition of the Collins Cobuild English Language Dictionary of 1987 (cf. also Moon 2007). Even if researchers choose not to adopt this “consultation paradigm” which starts from a prefabricated list of phraseological units but prefer an “analysis paradigm” (on these terms, cf. Sailer 2007: 1064) by retrieving hypothetical phraseological units from text corpora by various means, a database of accepted forms will finally be necessary for the evaluation of the potentially phraseological findings. As far as the “analysis paradigm” is concerned, a large amount of work on the available electronic corpora and their annotations will still be necessary before phraseological units which do not match the criterion of frequent lexical co-occurrence can be retrieved from them (cf., e.g., also Degand and Bestgen 2003 on the automatic retrieval of idioms). The problems and challenges of corpus-based phraseological study outlined by Sailer (2007: 1067–1069) for the analysis of Present-day German, for instance with regard to the variation in phraseological units, will have to be reviewed from the point of view of historical (English) corpora. Problems relating to the rare occurrence of many phraseological units and the comparatively small size of the databases are of course even more severe for historical than for present-day databases.
5 Origin, change, and earlier uses of English phraseological units Next to the identification of phraseological units in older texts, historical phraseological research has addressed questions of the origins and change of phraseological units through time. Many of these points are connected to key issues of historical linguistics. In addition, the uses of phraseological units in literary texts have been addressed, mainly with the aim to assess the literary language of individual authors or the style
187
188
II Linguistic Levels of particular text types in the past. These studies, however, also contribute to our knowledge of the forms, meanings and functions of phraseological units at particular points in time.
5.1 Origin and change of phraseological units as linguistic signs It seems safe to say that all phraseological units, once they are accepted parts of the language, share the feature of “conventionality”. It seems also safe to say that at the basis of all phraseological units lies the pragmatic feature of “usefulness”. These “useful units”, then, may develop from several origins (cf. Howarth 2000). Some idiomatic phraseological units may have started out from a free or restricted collocation, such as suggested by the semantic change of the complex predicate in (23). In these cases, classification criteria in typologies which have been suggested for the description of the rise of a phraseological unit are the structure from which the unit developed and the kind of semantic change of the basis compared to the result (cf. Munske 1993). Other phraseological units such as proverbs (4), oaths or the metaphorical phraseological unit in (24) are probably coined expressions with no original literal use. Some phraseological units which are in common use today originate from literary texts, that is, they can be traced back to one creative mind, as in (7). Borrowing, too, is another source of phraseological units – a source not to be underestimated for English as a mixed language – such as the foreign phrase in (25) or, more commonly, adapted or translated ones (26). However, one has to be aware that it can be difficult to distinguish borrowed units from common-source and polygenetic ones. (23) to take steps ‘walk’ > ‘perform a move or moves in a course of action’ (17th century) > ‘take action or measures towards attaining an end’ (from 18th century) (cf. also OED, s.v. step n.1, def. 6d) (24) to be in (another person’s) shoes ‘be in his position or place’ (from 18th century according to OED, s.v. shoe n., def. 2k) (25) vice versa ‘contrariwise, conversely’ (first attestation: 1601; cf. OED, s.v. vice versa adv.) (26) to keep company ‘give a person one’s company’ (first attestation: 1509; < tenir compaignie, according to Prins 1952) In cases when phraseological units develop from free collocations, the question of how such a phraseological “fixation” arises must be addressed. For the explanation of the rise of English adverbial phraseological units, for example, Jean Hudson (1998) has suggested a cyclical model of fixation which draws on discussions in cognitive linguistics, pragmatics, discourse analysis, grammaticalization theory, and other research in language change. Her model is unidirectional and circular. Put in a nutshell, Hudson
12 Linguistic Levels: Idioms and fixed expressions proposes that at the level of discourse, an ad hoc expression, that is, an expression formed on the principle of open choice, turns into a unit with a different (new?) meaning through pragmatic inferencing. This unit is not yet fixed. Due to this pragmatic inferencing, however, salience reduction of the constituent parts occurs on the second level, which is conceptualization. Hudson replaces the notion of “semantic decompositionality” by “cognitive analyzability”, which depends on salience. Reduced salience, in its turn, fixes the expression to some degree on the level of realization. Symptoms of fixedness are unexpected syntactic constraints on the constituent parts, such as the other day but *the other days, and unexpected collocational restrictions within the expression, e.g. first of all but *second of all. On a less theoretical level, signs of a fixation process have, among other things, been named by Burger and Linke (1998: 747–750) as the reduction of lexical variants, of possibilities of modification, and of variety in morphosyntactic structure (cf. also Forga´cs 2004 on aspects of lexicalization). With regard to the choice and establishment of a phraseological unit’s canonical form in the history of English, Voitl (1969) has tentatively suggested influence of prescriptive linguistics. However, the tension between convention and variation, as well as creative exploitation, can readily be seen as typical of living phraseological units. It is obviously a major challenge for historical phraseology to distinguish synchronic variation from diachronic change. The investigation of the formal and semantic change of English phraseological units, too, addresses key issues of historical linguistics. In Burger and Linke (1998) and Friedrich (2007), several distinctive changes are discussed. These changes affect, for example, the order of the elements within the unit (27), the reduction or increase in the number of elements (28), the exchange of lexemes (29), the change in syntactic structure (17), the change in meaning and use (23), and finally the death of a phraseological unit (20). While examples for changes of these kinds can partly be suggested by the entries in the OED and MED, these entries are, after all, highly selective and not intended to give precise phraseological details. In-depth studies focussing on English phraseological developments, based on a wide coverage of lexicographical sources and the use of a large corpus of historical texts, are wanting. (27) heels over head (from 14th century) ‘upside down’ > head over heels (from 18th century, with development of figurative sense; cf. OED, s.v. heel n.1, def. 15a, head n.1, def. 46b) (28) a bird in hand [is better than / worth two in the wood / bush] (15th –18th century) > a bird in the hand […] (from 15th century) (cf. OED, MED, other dictionary sources) (29) to know (best) where one’s shoe wrings (one) ‘know where (a person’s) difficulty or trouble is’ (from 14th century; last quotation 1887) > to know (best) where one’s shoe pinches (one) (from 16th century; cf. OED, s.v. pinch v., def. 5b, shoe n., def. 2f, wring v., def. 4b) Restrictions of space forbid a detailed review of further studies pertaining to the origin of phraseological units and their change over time. Therefore, a selection of the
189
190
II Linguistic Levels published ones will only be briefly mentioned, and one unpublished study will be referred to in some more detail. Thus, for example, foreign influence on the rise of English phraseological units is the subject of Prins (1952), who devoted a book-length study and several articles to the French Influence in English Phrasing. The traces left by Latin phrases in Old English were investigated by Gneuss (1955), and Gustaf Stern (1931: 224) in his model of semantic change included fixed expressions which are influenced by foreign languages. Several formal subtypes of phraseological units (see Section 2.2, above) that are of specific interest to English linguistics have become the object of study, too. Thus, the structural and semantic rise and historical development of phrasal verbs (e.g. to give up) and complex predicates (e.g. to take a bath) have been given a rather large amount of attention (cf. the monographs cited in Section 3, above). The problem of differentiating noun + noun compounds from syntactic groups with phraseological status has been tackled by Marchand (1969: 20–30), also in a historical perspective. Among studies investigating subtypes are Ross’s (1975) article on “alliterative phrases” and Sontheim’s (1972) monograph addressed to the historical development of English proverbs and proverbial phrases. Apart from individual formal subtypes, the potential of smaller phraseological units, such as the adverbial phraseological units in hand and on hand, for example, as elements in the formation of larger phraseological units (Fleischer’s 1997 phraseologische Reihe ‘phraseological chain’), such as in to take in hand or a bird in hand [is better than / worth two in the wood / bush] (28), for example, can fruitfully be explored in the history of English, too. An as yet unpublished study by the present author based on lexicographical sources has uncovered 62 phraseological units in which in hand and on hand appear in the history of English. Two of the results will be briefly mentioned here: First, a “phraseological semantic force” (my own term) of in hand and on hand can be found, for instance in the development of take in hand. All main semantic features of in hand seem to have gained control over the unit take in hand almost simultaneously until the feature [PROCESS] prevailed from the 16th century on. Thus, from the literal meaning ‘in the hand’ the feature [PROXIMITY] took hold (e.g. ‘to take with one’), also [POWER, CONTROL] ‘to bring under control’, and finally [PROCESS] ‘to carry out’, ‘to undertake’. The second conclusion relates to the “phraseological binding force” (my own term) of adverbial phraseological units such as in hand, which may vary over time. According to my data, most of the new units with in hand and on hand were formed in the 16th century, a period marked by unprecedented interest in the creative potential and rhetorical force of the mother tongue, aided above all by a desire for rhetorical copia, that is, full and variable English expression (on the latter aspect, cf. Knappe 2004: 49–111). The historical development of phraseological units can also be explored from an onomasiological point of view. Thus Anders (1995) has studied lexical and phraseological realizations of the concept of “dying”, also in a historical perspective. And finally, connecting this section up to Section 5.2, the historical change in form and function of English routine formulae can be studied. Thus, Wyld (1936) included as the tenth chapter of his history of modern colloquial English a discussion of the historical development of routine formulae and finds, for instance, that greetings and farewells in the 16th century were less “stereotypical” than in his own day. A useful synchronic collection and classification from which systematic studies of routine formulae
12 Linguistic Levels: Idioms and fixed expressions from a diachronic perspective could start is the monograph on Conversational Routines in English by Karin Aijmer (1996).
5.2 The use of phraseological units in historical texts: aspects of pragmatics Phraseological units have always been employed by writers in a variety of functions, and canonical poetry, vice versa, can itself act as the source of phraseological units, as in (7) (cf. also MacKenzie 2003). In addition to finding out what the employment of formulaic language means for the literary text, it can also give evidence on the (creative) use of phraseology at a given time in the history of a language. The stylistically motivated employment of phraseological units in literary texts has been emphasized in the Russian and (Eastern) European study of historical phraseology. English writers of the past have selectively been studied, too. To start at the beginning, it is well-known that the oral style in medieval poetry is characterized by formulaic expressions (cf. the classical article by Magoun 1953). These have received much attention in scholarship, and are examples for the construction and transmission of artistic texts by help of pre-set formulas, sometimes employed creatively. In later texts, the use of proverbs in particular has been investigated (cf. especially the bibliography on international proverb scholarship by Mieder 1982). To mention just a few examples, Weinstock (1966) looked at Shakespeare’s use of proverbs and found that in his later works, the great writer employed proverbs to mark dramatic turns, and they also served to characterize persons. Reuter (1986) studied Deloney from this perspective, and both Brewer (1986: 229–232) and Windeatt (1992: 332–335, 345–354) comment on Chaucer’s use of set phrases (e.g. bold as blind Bayard), oaths (e.g., God so my soule save) and proverbs (e.g. The blind man cannot judge in colors) in the discussion of his poetic style. Thus, for instance, one of Chaucer’s favourite lines (adapted from the Italian) is For pitee renneth soone in gentil herte ‘because pity flows swiftly in a noble heart’, which he uses in the Canterbury Tales and in the Legend of Good Women to reinforce “our sense of humane values”, according to Brewer (1986: 230). Brewer continues to say that in using a large number of familiar phrases Chaucer created a traditionalist diction and aimed at familiar effects, ready communication, and sympathetic attention. The sententiousness in Chaucer’s Troilus and Criseyde, particularly the use of proverbs, however, which cluster at important points of persuasion and self-persuasion, creates distinct ambiguity which questions the limitations and value of prudential wisdom (cf. Windeatt 1992: 345–354). These are only a few examples of employment of phraseological language in literary texts. Again, more work needs to be done along these lines.
6 The impact of phraseology and collocation / string frequency on language change By way of an outlook one further interesting line of research in historical English phraseology will be mentioned: this is the effect that multi-word combinations can have on language change. To this end, frequently repeated word combinations (frequent
191
192
II Linguistic Levels collocations) and phraseological units will be discussed together. First, the common ground of both concepts will be addressed. A description of a phraseological unit was given above (Section 2.1). The use of “collocation” as addressed here needs further explanation. As defined by Sinclair (1991: 170), collocation is “the occurrence of two or more words within a short space of each other in a text”. Typically, these co-occurrences are frequently repeated or statistically relevant in a corpus (cf. also Bartsch 2004: 65; Moon 2007: 1046–1047). Research into conceptual structures may help us understand the common ground of both collocations as described above and phraseological units in language processing and use, which is important for the question of their influence on language change. Both frequent word combinations and phraseological units are characterized by their degree of entrenchment (cf. Harris 1998): it is either based on the form only or on the semantic/pragmatic unity of a string with certain formal characteristics. The effect of frequently recurring word form combinations (string-frequency) on language change has received attention in scholarship, such as seen in coalescence phenomena and the phonological effect of word-boundary liaison as the trigger for the “Great Vowel Shift” (cf. Krug 1998 and 2003 and Chapter 48). So if the frequent co-occurrence of lexical items affects language change, as research in string frequency claims, it may be suggested that phraseological units, too, may have an impact on it. Thus, Knappe and Schu¨mann (2006) have investigated the influence of both collocations (as described above) and phraseological units on the sudden switches of pronouns of address between the singular and the plural in the address of a single person in dialogues in Chaucer’s Canterbury Tales. About 90 such switchings occur in the text. The use of the plural pronoun for formal address was introduced on the model of French but was by the time of Chaucer not yet vigorously applied. For about one third of the cases of sudden pronoun switch we claim that a “collocational-phraseological” force was the trigger. This force is so strong that it may even override the pragmatically preferred choice of the pronouns of address in the particular situation (cf. Jucker, Chapter 13). For instance, terms for body parts (e.g. thy eyen) tend to co-occur with the singular pronoun of address and may provoke a pronoun switch. The verb prayen ‘I pray you/thee’ occurs more often with the plural (62 instances) than with the singular pronoun (14 instances). To give an example, the more common combination of pray + you may account for the shift from the singular to the plural in Absolon’s speech to Alisoun in the Miller’s Tale (3361–3362), as seen in (30): (30) “Now, deere lady, if thy wille be, I praye yow that ye wole rewe on me …” ‘ “Now, dear lady, if it be your will, I beseech you to have mercy on me …” ’ In (31) the formula used for the opening of sermons Heere may ye se ‘By this may you see, i.e. understand’ is a phraseological unit. In the Friar’s Tale (1567) it can account for the insertion of the plural form in the yeoman’s (i.e. the devil’s) address to the summoner (1566–1568):
12 Linguistic Levels: Idioms and fixed expressions (31) “Lo, brother,” quod the feend, “what tolde I thee? Heere may ye se, myn owene deere brother, The carl spak oo thing, but he thoghte another.” ‘ “Lo, brother”, said the devil, “what did I tell you? By this may you see, my own dear brother, the fellow spoke one thing, but he thought another” ’ (The quotations in [30] and [31] are taken from Benson [ed.] 1987.) Starting from here, the role of entrenched structures in the introduction of a distinctive system of pronouns of address in Middle English could be studied.
7 Summary To sum up, a great amount of work in the retrieval and analysis of phraseological material from the history of English still needs to be done. Some first studies and research from other languages and in different traditions of scholarship can provide a basis from which to start with the systematic investigation of the phraseological past of the English language. In the long run, the aim of scholarship in historical English phraseology should be the description of all facets of the origin and historical development of the phraseological system, or systems, of English. Moreover, at a time when the impact of “formulaic language” on language acquisition, use, and change is increasingly being studied in linguistic theory, historical English phraseology in its interaction with other related approaches of scholarship and methods such as collocation studies, corpus linguistics, cognitive linguistics, grammaticalization and lexicalization theories, and also construction grammar, will open up large, promising fields of research possibilities.
8 References Aijmer, Karin. 1996. Conversational Routines in English: Convention and Creativity. London: Longman. Anders, Heidi. 1995. “Never say die”– Englische Idiome um den Tod und das Sterben. Frankfurt: Peter Lang. Apperson, George Latimer. 1929. English Proverbs and Proverbial Phrases: A Historical Dictionary. 2 vols. London/Toronto: J. M. Dent and Sons; New York: Dutton & Co. Bartsch, Sabine. 2004. Structural and Functional Properties of Collocations in English: A Corpus Study of Lexical and Pragmatic Constraints on Lexical Co-occurrence. Tu¨bingen: Narr. Benson, Larry D. (ed.). 1987. The Riverside Chaucer. 3rd edn. Oxford: Oxford University Press. Brewer, Derek. 1986. Chaucer’s poetic style. In: Piero Boitani and Jill Mann (eds.), The Cambridge Chaucer Companion, 227–242. Cambridge: Cambridge University Press. Brinton, Laurel J. 1996. Pragmatic Markers in English: Grammaticalization and Discourse Functions. Berlin/New York: Mouton de Gruyter. Brinton, Laurel and Minoji Akimoto (eds.). 1999. Collocational and Idiomatic Aspects of Composite Predicates in the History of English. Amsterdam/Philadelphia: John Benjamins. Burger, Harald. 2010. Phraseologie: Eine Einfu¨hrung am Beispiel des Deutschen. 4th edn. Berlin: Erich Schmidt. Burger, Harald, Annelies Buhofer, and Ambros Sialm. 1982. Handbuch der Phraseologie. Berlin/ New York: Mouton de Gruyter.
193
194
II Linguistic Levels Burger, Harald, Dmitrij Dobrovo 0skij, Peter Ku¨hn, and Neal R. Norrick (eds.). 2007. Phraseologie/ Phraseology: Ein internationales Handbuch der zeitgeno¨ssischen Forschung / An International Handbook of Contemporary Research. 2 vols. Berlin/New York: Mouton de Gruyter. Burger, Harald and Angelika Linke. 1998. Historische Phraseologie. In: Werner Besch, Anne Betten, Oskar Reichmann, and Stefan Sonderegger (eds.), Sprachgeschichte: Ein Handbuch zur Geschichte der deutschen Sprache und ihrer Erforschung. 2nd ed. Vol. I, 743–755. Berlin/ New York: Walter de Gruyter. Claridge, Claudia. 2000. Multi-word Verbs in Early Modern English: A Corpus Based Approach. Amsterdam/Atlanta, GA: Rodopi. Cowie, A. P. 1983. General introduction. In: Cowie, Mackin, and McCaig, Vol. II, x-xvii. Cowie, A. P. (ed.). 1998. Phraseology: Theory, Analysis, and Applications. Oxford: Clarendon Press. Cowie, A. P. 2003. Some aspects of the treatment of phraseology in the OED. In: Cornelia Tschichold (ed.), English Core Linguistics: Essays in Honour of D. J. Allerton, 205–224. Bern/Berlin: Peter Lang. Cowie, A. P., R. Mackin, and I. R. McCaig. 1983. Oxford Dictionary of Current Idiomatic English. Vol. II: Phrase, Clause & Sentence Idioms. Oxford: Oxford University Press. [Repr. 1993 under the title: Oxford Dictionary of English Idioms] Cowie, A. P. and R. Mackin. 1993. Oxford Dictionary of Phrasal Verbs. Oxford: Oxford University Press. [= 2nd edn. of Oxford Dictionary of Current Idiomatic English, Vol. I. Verbs with Prepositions & Particles] Degand, Liesbeth and Yves Bestgen. 2003. Towards automatic retrieval of idioms in French newspaper corpora. Literary and Linguistic Computing 18: 249–259. Dobrovo 0skij, Dimitrij, O. and Elisabeth Piirainen. 2009. Zur Theorie der Phaseologie: Kognitive and kulterelle Aspekte. Tu¨bingen: Staufferburg. Doyle, Charles Clay. 2007a. Historical phraseology of English. In: Burger, Dobrovo 0skij, Ku¨hn, and Norrick (eds.), 1078–1092. Doyle, Charles Clay. 2007b. Collections of proverbs and proverb dictionaries: Some historical observations on what’s in them and what’s not (with a note on current “gendered” proverbs). In: Skandera (ed.), 181–204. Fiedler, Sabine. 2007. English Phraseology: A Coursebook. Tu¨bingen: Narr. Fleischer, Wolfgang. 1997. Phraseologie der deutschen Gegenwartssprache. 2nd edn. Tu¨bingen: Niemeyer. Forga´cs, Tama´s. 2004. Grammatikalisierung und Lexikalisierung in phraseologischen Einheiten. In: Christine Palm-Meister (ed.), Europhras 2000: Internationale Tagung zur Phraseologie vom 15.–18. Juni 2000 in Aske / Schweden, 137–149. Tu¨bingen: Stauffenburg. Friedrich, Jesko. 2007. Historische Phraseologie des Deutschen. In: Burger, Dobrovo 0skij, Ku¨hn, and Norrick (eds.), 1092–1106. Gla¨ser, Rosemarie. 1986. Phraseologie der englischen Sprache. Leipzig: VEB Verlag Enzyklopa¨die. Tu¨bingen: Niemeyer. Gla¨ser, Rosemarie. 1998. The stylistic potential of phraseological units in the light of genre analysis. In: Cowie (ed.), 125–143. Gneuss, Helmut. 1955. Lehnbildungen und Lehnbedeutungen im Altenglischen. Berlin: Erich Schmidt. Ha¨cki Buhofer, Annelies and Harald Burger (eds.). 2006. Phraseology in Motion I: Methoden und Kritik. Akten der Internationalen Tagung zur Phraseologie (Basel, 2004). Baltmannsweiler: Schneider-Verlag Hohengehren. Harris, Catherine L. 1998. Psycholinguistic studies of entrenchment. In: Jean-Pierre Koenig (ed.), Conceptual Structure, Discourse and Language. Vol. II. Stanford, CA: CSLA. http://kybele. psych.cornell.edu/~edelman/Psych-426/Harris97-psycholinguistic-entrenchment.pdf Hiltunen, Risto. 1983. The Decline of the Prefixes and the Beginnings of the English Phrasal Verb: The Evidence from Some Old and Early Middle English Texts. Turku: Turun Yliopisto. Howarth, Peter. 2000. Describing diachronic change in English phraseology. In: Gloria Corpas Pastor (ed.), Las lenguas de Europa: Estudios de fraseologı´a, fraseografı´a y traduccio´n, 213– 230. Granada: Editorial Comares.
12 Linguistic Levels: Idioms and fixed expressions Huddleston, Rodney and Geoffrey K. Pullum. 2002. The Cambridge Grammar of the English Language. Cambridge: Cambridge University Press. Hudson, Jean. 1998. Perspectives on Fixedness: Applied and Theoretical. Lund: Lund University Press. Knappe, Gabriele. 2004. Idioms and Fixed Expressions in English Language Study before 1800: A Contribution to English Historical Phraseology. Frankfurt: Peter Lang. Knappe, Gabriele. 2006a. The treasury of phrases in Peter Mark Roget’s Thesaurus of English Words and Phrases (1852). In: Christoph Houswitschka, Gabriele Knappe, and Anja Mu¨ller (eds.), Anglistentag 2005 Bamberg: Proceedings, 475–487. Trier: Wissenschaftlicher Verlag. Knappe, Gabriele. 2006b. Peter Mark Roget’s Thesaurus of English Words and Phrases: A midnineteenth century example of the place of phraseology in the history of linguistic theory and practice. In: Christian Mair and Reinhard Heuberger (eds.), in collaboration with Josef Wallmannsberger, Corpora and the History of English: Papers Dedicated to Manfred Markus on the Occasion of His Sixty-Fifth Birthday, 205–220. Heidelberg: Winter. Knappe, Gabriele. 2006c. Phraseology in English language study before 1800 and Lewis Chambaud’s Idioms of the French and English Languages (1751). In: Ha¨cki Buhofer and Burger (eds.), 413–423. Knappe, Gabriele and Michael Schu¨mann. 2006. Thou and ye: A collocational-phraseological approach to pronoun change in Chaucer’s Canterbury Tales. Studia Anglica Posnaniensia 42: 213–238. http://ifa.amu.edu.pl/sap/files/42/17Knappe.pdf Krug, Manfred. 1998. String frequency: A cognitive motivating factor in coalescence, language processing and linguistic change. Journal of English Linguistics 26: 286–320. Krug, Manfred. 2003. (Great) Vowel Shifts present and past: Meeting ground for structural and natural phonologists. Penn Working Papers 9.2: Selected papers from NWAV 31: 107–122. Kurath, Hans, Sherman M. Kuhn, John Reidy, and Robert E. Lewis. 1952–2001. Middle English Dictionary. Ann Arbor: University of Michigan Press. http://quod.lib.umich.edu/m/med/ MacKenzie, Ian. 2003. Poetry and formulaic language. In: Christine Michaux and Marc Dominicy (eds.), Linguistic Approaches to Poetry, 75–86. Amsterdam/Philadelphia: John Benjamins. Magoun, Francis P. 1953. The oral-formulaic character of Anglo-Saxon narrative poetry. Speculum 28: 446–467. Makkai, Adam. 1972. Idiom Structure in English. The Hague/Paris: Mouton. Marchand, Hans. 1969. The Categories and Types of Present-Day English Word-Formation: A Synchronic-Diachronic Approach. 2nd edn. Munich: Beck. Mieder, Wolfgang. 1982. International Proverb Scholarship: An Annotated Bibliography. New York: Garland. Mieder, Wolfgang, Stewart A. Kingsbury, and Kelsie B. Harder (eds.). 1992. A Dictionary of American Proverbs. New York: Oxford University Press. Moon, Rosamund. 1998a. Fixed Expressions and Idioms in English: A Corpus-Based Approach. Oxford: Clarendon Press. Moon, Rosamund. 1998b. Frequencies and forms of phrasal lexemes in English. In: Cowie (ed.), 79–100. Moon, Rosamund. 2007. Corpus linguistic approaches with English corpora. In: Burger, Dobrovo 0skij, Ku¨hn, and Norrick (eds.), 1045–1059. Moralejo Ga´rate, Teresa. 2003. Composite Predicates in Middle English. Munich: LINCOM. Munske, Horst Haider. 1993. Wie entstehen Phraseologismen? In: Klaus J. Mattheier, KlausPeter Wegera, Walter Hoffmann, Ju¨rgen Macha, and Hans-Joachim Solms (eds.), Vielfalt des Deutschen, 481–515. Frankfurt am Main: Peter Lang. Nattinger, James R. and Jeanette S. DeCarrico. 1992. Lexical Phrases and Language Teaching. Oxford: Oxford University Press. Norrick, Neal R. 2007. English phraseology. In: Burger, Dobrovo 0skij, Ku¨hn, and Norrick (eds.), 615–619. Prins, A. A. 1952. French Influence in English Phrasing. Leiden: Universitaire Pers.
195
196
II Linguistic Levels Reuter, O. R. 1986. Proverbs, Proverbial Sentences and Phrases in Thomas Deloney’s Works. Helsinki: Societas Scientiarum Fennica. Ross, Alan S. C. 1975. “Run and Reve” and similar alliterative phrases. Neuphilologische Mitteilungen 76: 571–582. Sailer, Manfred. 2007. Corpus linguistic approaches with German corpora. In: Burger, Dobrovo 0skij, Ku¨hn, and Norrick (eds.), 1060–1071. Sialm, Ambros, Harald Burger, and Angelika Linke. 1982. Historische Phraseologie. In: Burger, Buhofer, and Sialm, 315–382. Sick, Christine. 1993. Adverbiale Phraseologismen des Englischen. Tu¨bingen: Narr. Simpson, John (ed.). 2000–. The Oxford English Dictionary. 3rd edn. online. Oxford University Press. www.oed.com/ Sinclair, John. 1991. Corpus, Concordance, Collocation. Oxford: Oxford University Press. Skandera, Paul (ed.). 2007. Phraseology and Culture in English. Berlin/New York: Mouton de Gruyter. Sontheim, Kurt. 1972. Sprichwort, sprichwo¨rtliche und metaphorische Redewendungen: Synchronische und diachronische Studien zu semantisch-idiomatischen Konstruktionen im Englischen. Ph.D. dissertation, University of Erlangen-Nu¨rnberg. Speake, Jennifer (ed.). 2003. The Oxford Dictionary of Proverbs. 4th edn. Oxford: Oxford University Press. Stern, Gustaf. 1931. Meaning and Change of Meaning: With Special Reference to the English Language. Westport, CT: Greenwood Press. Tilley, Morris Palmer. 1950. A Dictionary of the Proverbs in England in the Sixteenth and Seventeenth Centuries: A Collection of the Proverbs Found in English Literature and the Dictionaries of the Period. Ann Arbor: University of Michigan Press. Thun, Harald. 1978. Probleme der Phraseologie: Untersuchungen zur wiederholten Rede mit Beispielen aus dem Franzo¨sischen, Italienischen, Spanischen und Romanischen. Tu¨bingen: Niemeyer. Trousdale, Graeme. 2008. Constructions in grammaticalization and lexicalization: Evidence from the history of a composite predicate construction in English. In: Graeme Trousdale and Nikolas Gisborne (eds.), Constructional Approaches to English Grammar, 33–67. Berlin/New York: Mouton de Gruyter. Voitl, Herbert. 1969. Probleme der englischen Idiomatik. Germanisch-romanische Monatsschrift, N.F. 19: 194–212. Weickert, Rainer. 1997. Die Behandlung von Phraseologismen in ausgewa¨hlten Sprachlehren von Ickelsamer bis ins 19. Jahrhundert: Ein Beitrag zur historischen Phraseologie. Hamburg: Kovac. Weinstock, Horst. 1966. Die Funktion elisabethanischer Sprichwo¨rter und Pseudosprichwo¨rter bei Shakespeare. Heidelberg: Winter. Welte, Werner. 1992. On the properties of English phraseology: A critical survey. In: Claudia Blank (ed.), Language and Civilization: A Concerted Profusion of Essays and Studies in Honour of Otto Hietsch. Vol. II, 564–591. Frankfurt: Peter Lang. Whiting, Bartlett Jere, with the collaboration of Helen Wescott Whiting. 1968. Proverbs, Sentences, and Proverbial Phrases: From English Writings Mainly Before 1500. Cambridge, MA: Belknap Press of Harvard University Press. Windeatt, Barry. 1992. Troilus and Criseyde. Oxford: Oxford University Press. Wray, Alison. 2002. Formulaic Language and the Lexicon. Cambridge: Cambridge University Press. Wyld, Henry Cecil. 1936. A History of Modern Colloquial English. 3rd edn., with additions. Oxford: Basil Blackwell.
Gabriele Knappe, Bamberg (Germany)
13 Linguistic Levels: Pragmatics and discourse
197
13 Linguistic Levels: Pragmatics and discourse 1 2 3 4 5 6 7
Introduction Pragmatic explanations in language change Pragmatics as the study of performance phenomena Discourse as dialogue Discourse as a domain of communication Summary References
Abstract Pragmatics studies the processes of language use, while discourse analysis is devoted to its product, i.e. discourse. Pragmatics can be understood in a narrow sense focussing on cognitive-inferential aspects of information processing, and it can be understood in a wider sense in which it also includes social aspects of interaction. In historical pragmatics, the former conceptualization lies behind work on pragmatic explanations in language change, while the latter conceptualization studies earlier language use from a social and interactional perspective, including such aspects as inserts (e.g. interjections and discourse markers), speech acts, and terms of address. Discourse, as the product of language use, can be seen as a stretch of conversation (dialogue) or as a domain of communication. In the former conceptualization, research focuses on the structural properties of the dialogue, and in the latter, it deals with the linguistic practices pertaining to particular fields of knowledge or interaction, e.g. courtroom discourse, the discourse of science, and news discourse.
1 Introduction In a very general sense pragmatics can be defined as the study of language use, while discourse analysis, in an equally general sense, can be defined as the analysis of the result of human communication, viz. discourse. It has been suggested that discourse analysis is more text-centered, more static, more interested in product (in the well-formedness of texts), while pragmatics is more user-centred, more dynamic, more interested in the process of text production. Discourse analysis is frequently equated with conversational analysis, and pragmatics with speech act theory. It would seem difficult to distinguish the two with any conviction, however (Brinton 2001: 139).
There is certainly a great deal of overlap between the two fields. A large range of topics can be dealt with under either heading. Speech acts, such as greetings and farewells, or discourse markers, such as well, so, or you know have both interactional (pragmatic) functions and text-structuring or discourse functions. As a field of study, pragmatics has grown very considerably over the last thirty years or so. Traditionally, linguists were mainly concerned with an analysis of language structure at the levels of phonology, morphology, and syntax, but with the pragmatic turn in the late 1970s and early 1980s some of the interest shifted from the structure of Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 197–212
198
II Linguistic Levels language to the language user. At the beginning of this development, pragmatics was often seen as the ragbag of linguistic description (see Mey 1998: 716). As such it covered performance phenomena that could not be handled at the traditional levels of linguistic description, such as speech acts, conversational implicature, deixis, and politeness, but also the structure of conversations. On the other hand, even in the early days of pragmatics, the discipline was also seen as a perspective. As such it was not a level of linguistic description but a different way of analyzing language. Language was not seen as a system of signs but as a means of communication. “Pragmatics is a perspective on any aspect of language, at any level of structure” (Verschueren 1987: 5, italics in original; see also Verschueren 1999: 2). Under the former view, pragmatics was a separate level of linguistic description, parallel to other levels, such as syntax or semantics. Under the latter view, pragmatics was a particular way of doing linguistics that could be applied to all other levels of linguistic description from phonology and morphology to syntax, semantics and, indeed, discourse. These positions have developed into a more restricted cognitive-inferential conceptualization of pragmatics (adhered to, generally speaking, by Anglo-American researchers) and a broader socio-interactional conceptualization (common among European researchers). Cruse (2000), for instance, gives the following narrow definition of pragmatics: For present purposes pragmatics can be taken to be concerned with aspects of information (in the widest sense) conveyed through language which (a) are not encoded by generally accepted convention in the linguistic forms used, but which (b) none the less arise naturally out of and depend on the meanings conventionally encoded in the linguistic forms used, taken in conjunction with the context in which the forms are used (Cruse 2000: 16).
In this conceptualization, people routinely understand more than what is explicitly communicated. They read between the lines, as it were, and this is the field of the pragmaticist. In her handbook article on historical pragmatics, Traugott (2004: 539) also takes pragmatics “to be non-literal meaning that arises in language use”, and Sperber and Noveck (2004: 1) define pragmatics as “the study of how linguistic properties and contextual factors interact in the interpretation of utterances.” In their view, pragmatics is not restricted to a study of implicit meanings. In fact, they are at pains to demonstrate that there are many aspects of explicit meaning that require access to contextual information for their interpretation, but they exclude the wider social issues of language use from the scope of pragmatics. The European tradition adopts a broader, more sociologically based view of pragmatics that includes social and cultural conditions of language use. Trosborg (1994: 37), a representative of this broader European tradition, for instance, states that “sociopragmatics is concerned with the analysis of significant patterns of interaction in particular social situations and/or in particular social systems. For example, speech acts may be realized differently in different social contexts and situations as well as in different social groups within a speech community,” while Blakemore, a representative of the Anglo-American tradition, finds it “misleading to include phenomena like politeness, face-saving and turn taking […] under the general heading of pragmatics” (Blakemore 1992: 47). The two conceptualizations of pragmatics, obviously, have consequences for the interaction of pragmatics and historical linguistics. The former conceptualization
13 Linguistic Levels: Pragmatics and discourse suggests a range of specific performance-related topics, while the latter suggests a specific way of investigating earlier stages of a language and its development. The term “discourse” is perhaps even more open to different definitions. On the one hand, it can be seen as the spoken equivalent of a text. A (written) text is made up of sentences while a (spoken) discourse is made up of utterances. In this sense, the term “discourse” is more or less synonymous with the term “dialogue” (see below, Section 4). Brinton (2001: 139–140) distinguishes between three discourse analytical approaches to historical data. First, the discourse analyst may use forms, functions, and structures of discourse at historical stages of a language. She calls this approach “historical discourse analysis proper”. Second, the discourse analyst may study the discourse-pragmatic factors and motivations behind language change. This approach is called “discourseoriented historical linguistics”. And third, the discourse analyst may focus on the diachronic development of discourse functions and discourse structures over time. She calls this third approach “diachronic(ally oriented) discourse analysis”. However, the term “discourse” can also be used in a much wider sense, not just for a linguistic unit larger than utterances, but as a domain of language. In such a view, a discourse is a collection of linguistic practices characterized by a distinct group of people and a distinct group of genres and text types, e.g. the discourse of science, or more specifically the discourse of medical science or the discourse of modern linguistics. In the following I shall evaluate how these conceptualizations of the terms “pragmatics” and “discourse” can be applied to the analysis of historical data and in particular to English historical data.
2 Pragmatic explanations in language change In the Anglo-American conceptualization of pragmatics, pragmatics is mainly a tool to describe and explain patterns of language change. Language is a means of communication and, therefore, the communicative forces that are at work when people use language must be taken into consideration when we analyse, for instance, the syntax of a language and indeed when we analyse diachronic changes in the syntax of a language. Thus, pragmatics becomes a principle of explanation in language change. In Brinton’s (2001) terminology this would be “discourse-oriented historical linguistics”. If pragmatics is seen as one level of linguistic description on a par with other levels such as phonology, morphology, syntax and semantics, it is largely restricted to nontruth-conditional aspects of language, and to aspects of language that depend on the context of utterance. Deictic elements, for instance, depend on the situation of use for their interpretation. Speech acts in their early conceptualization of doing things with words were also restricted to non-truth-conditional aspects. Speech act theory took its starting point from Austin’s (1962) observation that speech acts are regularly used for purposes other than stating facts that are assessable in terms of true or false. Meanings are not abstract entities that pertain to linguistic expressions but the result of negotiations between speaker/writer and addressee/reader, which – through repetition of use – have become conventionalized. A theory of meaning change, therefore, must take into account the communicative situation of speaker/writer and addressee/ reader. Traugott and Dasher (2005), for instance, argue that it is ad-hoc negotiations of meanings that may lead to meaning change if they are invoked repeatedly until they become conventionalized in the entire speech community. They call such ad-hoc
199
200
II Linguistic Levels meanings “invited inferences”, a term borrowed from Geis and Zwicky (1971). However, Traugott and Dasher use it in a broader sense and do not restrict it to generalized implicatures. It signals the speaker/writer’s role in inviting the addressee to infer the intended ad-hoc meaning. As an example they cite the case of as/so long as (Traugott and Dasher 2005: 36–37). In Old and Middle English the spatial meaning (‘of the same length as’) co-existed with the temporal meaning (‘for the same length of time as’). In some contexts, the meaning invited the conditional meaning ‘provided that’, as for instance in (1). (1)
wring þurh linenne clað on þæt eage swa lange swa him ðearf sy. (850–950 Lacnunga, p. 100; Traugott and Dasher 2005: 36, ex. 19) wring through linen cloth on that eye as long as him need be-SUBJ ‘squeeze (the medication) through a linen cloth onto the eye as long as he needs.’
The medicine is to be applied for the duration that it is needed, which invites the inference that it is to be applied only if it is needed. According to Traugott and Dasher all examples of as/so long as in Old and Middle English are either spatial or temporal, and while some allow a conditional reading, the conditional reading is never predominant. This changes in Early Modern English, when examples occur in which the invited inference of conditionality has been generalized to contexts of reasoning and cognition in which a temporal reading does not make sense or is at least not salient as in (2). (2)
They whose words doe most shew forth their wise vnderstanding, and whose lips doe vtter the purest knowledge, so as long as they vnderstand and speake as men, are they not faine sundry waies to excuse themselues? (1614 Hooker, p. 5; Traugott and Dasher 2005: 37, ex. 20)
Here the conditional reading is salient, while the temporal meaning is still available. Traugott and Dasher paraphrase the temporal meaning as “for the time that they understand and speak as men”, i.e. “as long as they live”. From the mid-19th century there are examples in which the conditional is the only possible meaning as in (3). (3)
“Would you tell me, please, which way I ought to go from here?” “That depends a good deal on where you want to get to,” said the Cat. “I don’t much care where-” said Alice. “Then it doesn’t matter which way you go,” said the Cat. “- so long as I get somewhere,” Alice added as an explanation. (1865 Carroll, Chapter 6, p. 51; Traugott and Dasher 2005: 37, ex. 21a)
Thus meaning change is the result of the interaction between speakers/writers and addressees/hearers in communicative situations. Speakers/writers use established coded meanings (e.g. the temporal reading of so/as long as) in creative ways to invite inferences. Through repeated use, such invited inferences become conventionalized and ultimately they become new coded meanings (Traugott and Dasher 2005: 38). Thus language change is seen as the result of what Keller (1994) has called an “invisible hand process”. Language change comes about as a causal effect of the accumulation of individual speakers’ action, who – individually – did not intend this effect.
13 Linguistic Levels: Pragmatics and discourse
3 Pragmatics as the study of performance phenomena Performance phenomena pertain mostly to the spoken language, i.e. to language that is produced under the constraints of online production. Such phenomena were shunned as irrelevant for a long time. For historical linguists they were doubly irrelevant. They were irrelevant because they were not part of the language system itself, and they were irrelevant because historical linguists did not have access to the spoken language of the past. The communicative turn in the ’70s and ’80s of the 20th century turned performance phenomena into legitimate objects of investigation for synchronic linguistics. Pragmaticists focused their attention on transcriptions of spoken interaction. They studied the minutiae of the turn-taking system, the form and function of individual utterances (speech acts), and so on. But these studies were restricted to present-day data. Pragmaticists saw written language as secondary and therefore as uninteresting for pragmatic analyses. Today performance phenomena have made their way into standard descriptions of the English language (e.g. Biber et al. 1999, who spend a considerable amount of space on such phenomena within the confines of a structural description of the English language), and within the last decade or so, significant progress has been made on the description of performance phenomena from a diachronic perspective. I shall briefly mention three examples which have received a considerable amount of attention from historical pragmaticists, inserts, speech acts and terms of address. To the extent that the analyses of these elements rely on references to social conditions of their use, they clearly go beyond the narrow Anglo-American conceptualization of pragmatics.
3.1 Inserts Biber et al. (1999: 1082) use the term “inserts” to refer to “stand-alone words which are characterized in general by their inability to enter into syntactic relations with other structures. […] They comprise a class of words that is peripheral, both in the grammar and in the lexicon of the language”. They distinguish nine different types of inserts: interjections (oh, ah), greetings and farewells (hi, hello, goodbye), discourse markers (well, right), attention signals (hey, yo), response elicitors (right?, eh?), response forms (yeah, yep), hesitators (um, er), various polite speech-act formulae (thanks, sorry), and expletives (shit, good grief!). Not all of these are equally amenable to a historical analysis. Biber et al. (1999: 1096–1098) provide some statistics about their distribution in American English and British English conversations, but they do not say anything about their occurrence in written genres. It seems reasonable to assume that some of them are relatively infrequent in the texts that have survived from earlier centuries. While some inserts, such as interjections or discourse markers, have been analyzed in their own right, others, like thanks and sorry, have been investigated in larger contexts of speech act studies of thanking and apologizing (e.g. Jacobsson 2002; Jucker and Taavitsainen 2008b), and expletives have been investigated in the context of the language of insults (e.g. Craun 1997). Taavitsainen (1995) investigates the form, function and distribution of exclamations, such as alas, ey, ah, harrow, and O in Late Middle and Early Modern English (see also Hiltunen 2006; Person 2009). Their distribution is clearly genre specific. In the Helsinki
201
202
II Linguistic Levels Corpus, which was used for the investigation, exclamations were particularly frequent in the genres comedy and fiction. They also occurred in trials and in Bible texts. In other genres they were rare. Exclamations were used more widely and with a broader variety of functions than in Present-day English. They were regularly used as vocatives and as appeals to the addressee. The interjection O, for instance, is often prefixed to an exclamatory sentence and it often combines with a vocative as in example (4), which is taken from a sermon. (4)
O my God, my God, why haste thou forsaken me? (1614 Hooker, Two Sermons Upon Part of S. Judes Epistle, 1614, p. 7; Helsinki Corpus, Taavitsainen 1995: 453)
Discourse markers have received considerable attention in historical pragmatics. Brinton (1996), for instance, analyzed a broad range of discourse markers, or “pragmatic markers”, as she calls them, including Old English hwæt, Middle English gan, and Middle and Early Modern English anon. She is interested not only in the developing discourse functions of these elements but also in the grammaticalization processes that they instantiate. In more recent publications she has added analyses of only (Brinton 1998), I say (Brinton 2005) and I mean (Brinton 2007) (see also Jucker 1997, 2002; Fischer 1998; Brinton 2006).
3.2 Speech acts Speech acts are not easily amenable to historical investigations because the traditional research methods developed for present-day languages cannot be applied to historical data. Originally the concept was developed by philosophers who investigated the nature of speech acts on the basis of careful considerations of what it means to name a ship, to make a promise, to issue a command, to ask a question, or to greet somebody (Austin 1962; Searle 1969). Later, empirical methods, such as discourse completion tests and role-plays, were developed to investigate speech acts and their realisations by different groups of speakers (e.g. Blum-Kulka et al. 1989; Trosborg 1994). For obvious reasons, none of these methods can be applied to historical data. More recently, corpus-based research methods have been improved and developed to such an extent that various avenues of investigations of historical speech act material have become available. It is, of course, possible to search for verbs denoting specific speech acts. Such speech act verbs are sometimes used performatively to carry out the speech act they denote. Kohnen (2008a), for instance, argues that in Old English explicit performatives were typically used to issue requests and commands as in (5): (5)
Ic bidde eow þæt ʒe ʒymon eowra sylfra, swa eowere bec eow wissiað. (Ælfric, Letter to Wulfsige, 26; Helsinki Corpus, Kohnen 2008a: 30) ‘I ask you to take care of yourselves, as your books teach you.’
The Old English verb biddan ‘ask, bid’ is here used performatively. By saying Ic bidde eow ‘I ask you’ the speaker carries out the speech act of asking or requesting (see in particular Kohnen 2000). However, many verbs that describe a speech act are not normally used performatively. They are used to talk about the speech act they name. They may occur in
13 Linguistic Levels: Pragmatics and discourse narratives with an account that a particular speech act had been performed, or in negotiations when the precise speech act value of an utterance is being discussed. (6)
If eny man wolde challenge a frere of Seint Frauncessis ordre and seue … Frere, thou louest money as myche as othere men […] (c.1449 Pecock Repr.; Taavitsainen and Jucker 2007: 113) ‘If any man were to challenge a friar of the order of St. Francis and to say … “Friar, you love money as much as other men […]’
In (6) the speech act verb “challenge” is used together with an example of an utterance with this speech act value. Many speech acts, perhaps most, are carried out without the relevant speech act verb. In order to locate relevant speech acts, the researcher has to rely on the philological method of actually reading the source texts. Jucker and Taavitsainen (2000) have used this method to describe insults in the history of English. But the method obviously precludes any statistical results. The findings can only be very selective based on the available research time. Some speech acts show recurrent surface patterns. Deutschmann (2003), for instance, has shown that apologies in English are mostly formulaic. They can be traced with corpus-linguistic tools by searching for a small number of expressions that typically occur in apologies, such as sorry, pardon, and excuse together with related and expanded forms. The same method has recently been used to trace apologies (Jucker and Taavitsainen 2008b), promises (Valkonen 2008) and compliments (Jucker et al. 2008).
3.3 Terms of address In the 13th century under the influence from French, English started to use the second person plural pronoun ye not only for two or more addressees but – under certain circumstances – also for one single addressee. Many Indo-European languages still have this distinction between two pronominal forms of address for a single addressee. On the basis of Latin tu and vos, the pronoun choices are usually abbreviated as T and V (Brown and Gilman 1960: 254). The conditions under which one pronoun or the other is chosen have been the object of extensive research in recent years (see, for instance, the volume by Taavitsainen and Jucker 2003). Brown and Gilman (1960) in their seminal article on the topic tried to find a common denominator for all languages with such a system. They argue that this common denominator is the semantics of power and solidarity. In medieval Europe, according to this theory, the power semantics accounted for a non-reciprocal use of T from the more powerful to the less powerful. The more powerful received V in return from the less powerful. Equals of the upper classes exchanged mutual V, while equals of the lower social classes exchanged mutual T. The power semantics of medieval Europe has been replaced by the solidarity semantics in which mutual V signals distance and mutual T solidarity. A significant body of research has shown that social conditions for the choice of T or V in specific situations are considerably more complex. Mazzon (2000), Honegger (2003), and Jucker (2006), for instance, have shown that Chaucer’s system of pronoun choices is much more situationally governed than the usual present-day systems in languages such as German, French or Italian. In the present-day forms of these languages,
203
204
II Linguistic Levels choices are more or less fixed for any given dyad of speakers, and a switch from mutual V to mutual T is a noticeable event, often accompanied by some kind of ritual (a switch from mutual T to mutual V, i.e. from informal to formal, would be very unusual). In Chaucer’s English, the characters of his fictional work used a more complex system that was based not only on social status between the characters but also on the basis of situational dominance or subjugation. Such approaches have replaced the earlier accounts of Chaucer’s use of personal pronouns by such scholars as Nathan (1959), Wilcockson (1980), and Burnley (1983), who tried to explain the choices largely on the basis of fixed social relationships. By the time of Shakespeare, it does no longer seem possible to provide an account that explains individual pronoun choices. Researchers, therefore, generally focus on frequencies and on co-occurrence patterns of nominal and pronominal terms of address. U. Busse (2002, 2003), for instance, shows that titles of courtesy, such as Your Grace, Your Ladyship, (my) liege or sir, are more likely to occur together with a V pronoun than any of the other categories of nominal terms of address, while terms of endearment, such as bully, chuck, heart, joy or love are most likely to occur together with a T pronoun (see also Stein 2003; B. Busse 2006).
4 Discourse as dialogue Discourse can be seen as a stretch of conversation or as a domain of language. In this section, I will use the term “dialogue” to refer to the former and the term “domain of discourse” for the latter. The terms “discourse” and “dialogue” imply an interaction between a speaker or writer and a recipient. Written texts, although there is no regular exchange of roles between speaker/writer and hearer, do have an addressee, even if the addressee is only a recipient and cannot actively contribute to the interaction. They are what Kilian (2005: 102) identifies as a “functional” dialogue. Fritz (1995: 469) distinguishes three stages of what he calls “historical dialogue analysis”. The first stage is characterized by analysis of the pragmatic structure and function of a historical dialogue in its social and historical context. The second stage is characterized by a contrastive comparison of earlier dialogue forms with later dialogue forms. The third and most advanced stage is characterized by an investigation of the evolution and dissemination of specific forms of dialogue. In the first stage, the researcher can use the same conversation analytical or dialogue analytical tools that are employed in modern data in order to investigate older forms of dialogue. The analysis can either adopt a macro perspective or a micro perspective. Under the macro perspective, the researcher focuses on the structure of the dialogue under analysis. Levinson (1983) reserved the term “discourse analysis” for such macro analyses of dialogue structures. Under the micro perspective, the researcher focuses on individual pragmatic elements, such as greetings, address terms, discourse markers and so on; or on local structures, e.g. adjacency pairs, such as question – answer sequences. Levinson (1983) used the term “conversation analysis” for this type of investigation. An analysis of individual pragmatic elements in individual dialogues of earlier periods coincides with the pragmatic research interests sketched out above. And indeed, a considerable amount of research has been published, e.g. on address terms in Chaucer’s narratives or in Shakespeare’s plays (see Section 3.3). But researchers have also
13 Linguistic Levels: Pragmatics and discourse adopted the larger perspective of looking at the inventory of pragmatic elements making up a specific type of historical dialogue. Watts (1999), for instance, investigates in detail two dialogues that were printed in 16th-century English language coursebooks for the benefit of learners of English as a foreign language. However, in practice it is not always easy to distinguish between the different stages envisaged by Fritz. Jucker and Taavitsainen (2000), for instance, investigate the use of insults in the history of English. The aim is to show a development or an evolution from the earlier forms to the later forms, but at present all that seems to be possible is a contrastive analysis of selected examples at different periods in the history of English. It is not yet possible to trace a continuous evolution of specific speech acts, such as insults. Archer in various publications (e.g. Archer 2005, 2006, 2007) gives a detailed picture of Early Modern English courtroom dialogue and thus carries out research at the first stage of historical dialogue analysis, but she also compares these findings to the present-day courtroom, representing the second stage. And finally she also draws attention to developments within the period under investigation, and thus contributes to stage three of historical dialogue analysis. She focuses mainly on the question-answer sequences in the courtroom dialogues and uses these to pinpoint the (changing) discursive roles of the active participants in the English courtroom, i.e. the judges, lawyers, witnesses and defendants. Taavitsainen (1999) also investigates the evolution of a particular form of dialogue. She assesses medical dialogues in Late Middle and Early Modern English, and traces the evolution of these dialogues between 1375 and 1750. She describes two traditions that are evident in Early English medical dialogues: the scholastic formula, based on the format of debates by Greek philosophers, and the mimetic dialogues, in which material is presented in fictional conversations between the author and the reader or between fictional characters. Taavitsainen shows how these traditions develop over the centuries and how, in the 18th century, medical dialogues merge with the new pamphlet tradition, in which social matters, such as health-care for the poor or polite conversations, are treated.
5 Discourse as a domain of communication As pointed out in Section 1, the term “discourse” can also be used in a more general sense as the totality of linguistic practices that pertain to a particular field of knowledge or to a particular occupation. Such discourses consist not of utterances but of typical text types, characterized by specific lexical items, idiosyncratic syntax, and particular routinized patterns of interaction. In such a context, researchers also ask more general questions about the dissemination of information within groups of speakers. Three such domains of communication in particular have received a fair amount of scholarly attention for the Early Modern English period: courtroom discourse, the discourse of science and news discourse.
5.1 Courtroom discourse A considerable amount of research has appeared on courtroom discourse in the Early Modern English period. The Early Modern English courtroom differed considerably from its modern equivalent. While modern courts presume a defendant to be innocent
205
206
II Linguistic Levels until proven guilty, the Early Modern courtroom expected the defendants to prove their innocence. Archer (2005: 85) demonstrates how this leads to a more active involvement on the part of the defendant. It was only in the later part of the Early Modern period that courtrooms introduced defence counsels who started to speak on behalf of the defendant. Koch (1999: 410–411), in his analysis of excerpts of three early Romance court records, draws attention to the communicative complexity of such records. The records written by a court scribe and addressed to a future reader are legal documents with appropriate formality of expression especially in the ritualistic elements pertaining to the formalities of the proceedings. These parts of the court records are characterized by the “language of distance” as Koch calls it. Embedded in this formal document there is a transcription of the verbal interaction taking place in the courtroom between the judge, the witnesses, the defendants and the lawyers. These utterances, even if they are written down, are closer to spoken language, or the “language of immediacy”. There may even be further embeddings, especially if the court cases dealt with libel, in which courtroom interactants report utterances that were spoken outside the courtroom. Such reported utterances are even closer to the language of immediacy. In her work on the Early Modern English courtroom Archer (2005, 2006, 2007) draws a detailed picture of the strategies adopted by the judge, the lawyers, the defendants and the witnesses. She concludes that the frequency of questions, their function and their interactional success depended on a number of sociopragmatic factors, such as the speech event, the position of the question and the discursive roles of the speaker and the addressee as well as the date of the trial (2005: 281). Culpeper and Semino (2000) extend the scope of courtroom discourse. They use two types of data, learned treatises on the topic of witchcraft and courtroom witness depositions. In their analysis, they deal with speech act verbs, such as to curse and they show how such verbs could be used to reinterpret trivial arguments within a village community into a witchcraft event. The witch trials that took place in 1692 in the Puritan village of Salem in the colony of Massachusetts have attracted a considerable amount of research into the discourse strategies adopted by the participants and the functional and structural properties of the trials as such. Kahlas-Tarkka and Rissanen (2007), for instance, investigated the discourse strategies of “successful” and “unsuccessful” defendants in the Salem witch trials, while Hiltunen and Peikola (2007) focus on the material evidence of these trials, i.e. the handwritten records and the printed editions. Their contribution demonstrates vividly how important it is not to forget the communicative role of the scribe who commits the spoken words in the courtroom to writing and thus makes it available for future generations (see also Doty and Hiltunen 2002; Hiltunen 2004; Doty 2007).
5.2 The discourse of science In the late medieval world, the discourse of science was multilingual. The main language for written texts was Latin, but texts started to be translated into the vernacular and the Greco-Roman tradition provided a model for scientific writing in the vernacular. In modern linguistics, “medical discourse” refers collectively to the communicative practices of the medical profession, both written and spoken. In the late medieval period, the medical profession consisted of heterogeneous groups of practitioners, including
13 Linguistic Levels: Pragmatics and discourse physicians, surgeons, barbers, midwives, itinerant specialists (e.g., bonesetters and oculists), herbalists, apothecaries, wisewomen, and others. They can be roughly divided into clerical and elite practitioners and tradespeople or ordinary practitioners; literacy was restricted mostly to the elite group (Taavitsainen 2006: 688).
Taavitsainen (2006) gives an overview of genres that were important for this discourse community. Compilations and commentaries of earlier studies were important for the dissemination of scholastic knowledge. Texts in question-and-answer format and pedagogical dialogues were also popular genres of scientific and medical writing that were adopted from Latin models into the vernacular. The volume edited by Taavitsainen and Pahta (2004) contains a range of detailed studies of medical and scientific writing in Late Medieval English. Ma¨kinen (2004), for instance, describes Middle English herbal recipes and recipes in manuals for medicinal plants and shows the textual traditions that link them together. Valle (1999: vii) takes the view that “science has at least since the seventeenth century taken place within a knowledge-producing discourse community, and that this community will in some way be ‘represented’ in scientific texts, in forms which can be identified and studied”. The totality of texts produced by this discourse community is, therefore, the discourse of science. In her study, Valle describes the discourse community of the Royal Society on the basis of a corpus of texts drawn from the Philosophical Transaction, spanning the three centuries from the beginning of publication in 1665 to 1965 (see also Valle 1997, 2006). Gotti (2006), too, deals with the discourse community of the Royal Society in London and illustrates some of the methods that were used by this community to spread the news about new discoveries and other scientific findings. Letters exchanged between scholars played an important role. They were not only exchanged between individuals but they were frequently copied and passed on to new recipients. Some influential scholars at the centre of scientific networks regularly received, sent and resent a large number of letters and thus had the role of clearing houses.
5.3 Early English news discourse With the invention of the printing press it became possible to publish accounts of recent events and to disseminate them to a large audience. In the 16th and 17th centuries pamphlets and newsbooks were used for this purpose (Raymond 2003). The first newspapers in the modern sense appeared in the early 17th century, first on the continent but soon also in England (Brownlees 1999; Studer 2008). The first newspapers or corantos, as they were originally called, consisted mainly of dispatches from correspondents from important places throughout Europe. These letters were inserted into the newspaper in the order in which they arrived at the editorial office in London. There was no other structural principle. It took another century for the first daily newspapers to be published in the early 18th century. As Sommerville (1996) has pointed out, the revolutionary aspect of this kind of news discourse consisted in the fact that newspapers appeared in regular intervals, weekly at first, twice or three times a week later and then daily. Thus, news was no longer reported in response to important events but a certain amount of space had to be filled with news on a regular basis. The early news discourse has attracted a fair amount of research recently not only in collections of articles, such as Ungerer (2000), Herring (2003), Raymond (2006) or
207
208
II Linguistic Levels Brownlees (2006) but also in monographs. Studer (2008), for instance, develops a larger picture of the development of news discourse on the basis of the Zurich English Newspaper Corpus (ZEN). He argues that news discourse is shaped by such external factors as the historical context and technological innovations. News discourse both adopted and adapted generic conventions; that is to say, it used existing genres, e.g. in the form of the letters from correspondents in the early newspapers, and it transformed and shaped them for its own needs. (See further Fries, Chapter 67.)
6 Summary It is not possible to draw a principled distinction between historical topics that are treated with pragmatic tools of investigation and those that are treated with discourse analytical tools. Traditionally, those approaches that focus on the interactional and dynamic aspects of language belong to pragmatics while those that focus on the structural aspects of dialogues, conversations or discourses belong to discourse analysis. The application of pragmatic and discourse analytical tools to historical data has uncovered a rich area of investigation and thrown new light on much familiar data. But a lot still needs to be done. At present, three areas of research appear to be particularly promising. First, the research on the history of speech acts has only just started to attract more than just occasional research efforts. In the volume edited by Jucker and Taavitsainen (2008a) a number of researchers have joined forces to investigate a range of different speech acts in the history of English and to develop the necessary methodologies. Recent advances in corpus technology have made it increasingly possible to locate some speech acts automatically. Second, the research of the evolution of forms of dialogue is still in its infancy. Kilian (2005) has presented an introduction into historical dialogue research, in which he develops a detailed typology of historical types of dialogues and some methodologies to investigate a broad range of such dialogues, i.e. dialogues in which speakers and addressees take turns in their roles. Culpeper and Kyto¨ (2010: 2) ask: “what was the spoken face-to-face interaction of past periods like?” in a systematic way and approach this question from various angles. In particular they look at the structure of conversations, at what they call “pragmatic noise”, i.e. pragmatic interjections or discourse markers, and social roles and gender in interaction. And third, the evolution of domains of discourse appears to be a very promising field of research. The existing work on courtroom discourse, the discourse of science and news discourse needs to be continued, and other domains should be tackled. The discourse of religion, for instance, would be an obvious candidate because there is wealth of historical material available consisting of many different text types, such as sermons, prayers, treatises and saints’ lives. The compilation at the University of Cologne of a Corpus of English Religious Prose is very likely to be a first significant step in this direction (see Kohnen 2007). Thus it seems that the new corpora and advances in corpus linguistics have had and are having a considerable impact on historical pragmatics and historical discourse analysis. The cooperation between corpus linguists and historical pragmaticists/discourse analysts has only just started, but it promises considerable advances in our understanding of human interaction and communication from a historical perspective.
13 Linguistic Levels: Pragmatics and discourse Acknowledgments: I thank Thomas Kohnen, Daniela Landert, and Elizabeth C. Traugott for valuable comments on a draft version of this paper. The usual disclaimers apply.
7 References Archer, Dawn. 2005. Questions and Answers in the English Courtroom (1640–1760). Amsterdam/ Philadelphia: John Benjamins. Archer, Dawn. 2006. (Re)Initiating strategies: Judges and defendants in Early Modern English courtrooms. Journal of Historical Pragmatics (Special Issue on Historical Courtroom Discourse, ed. by Barbara Kryk-Kastovsky) 7(2):181–211. Archer, Dawn. 2007. Developing a more detailed picture of the English courtroom (1640–1760): Data and methodological issues facing historical pragmatics. In: Fitzmaurice and Taavitsainen (eds.), 185–217. Austin, J. L. 1962. How to Do Things With Words. Oxford: Oxford University Press. Biber, Douglas, Stig Johansson, Geoffrey Leech, Susan Conrad, and Edward Finegan. 1999. Longman Grammar of Spoken and Written English. London: Longman. Blakemore, Diane. 1992. Understanding Utterances. An Introduction to Pragmatics. Oxford: Blackwell. Blum-Kulka, Shoshana, Juliane House, and Gabriele Kasper (eds.). 1989. Cross-Cultural Pragmatics: Requests and Apologies. Norwood, NJ: Ablex. Borgmeier, Raimund, Herbert Grabes, and Andreas H. Jucker (eds.). 1998. Anglistentag 1997 Giessen. Proceedings. Trier: Wissenschaftlicher Verlag. Brinton, Laurel J. 1996. Pragmatic Markers in English. Grammaticalization and Discourse Functions. Berlin/New York: Mouton de Gruyter. Brinton, Laurel J. 1998. “The flowers are lovely; only, they have no scent.”: The evolution of a pragmatic marker in English. In: Borgmeier, Grabes, and Jucker (eds.), 9–33. Brinton, Laurel J. 2001. Historical discourse analysis. In: Deborah Schiffrin, Deborah Tannen, and Heidi E. Hamilton (eds.), The Handbook of Discourse Analysis, 138–160. Oxford: Blackwell. Brinton, Laurel J. 2005. Processes underlying the development of pragmatic markers: The case of (I) say. In: Janne Skaffari, Matti Peikola, Ruth Carroll, Risto Hiltunen, and Brita Wa˚rvik (eds.), Opening Windows on Texts and Discourses of the Past, 279–299. Amsterdam/Philadelphia: John Benjamins. Brinton, Laurel. 2006. Pathways in the development of pragmatic markers in English. In: Ans van Kemenade and Bettelou Los (eds.), The Handbook of the History of English, 307–334. Oxford: Blackwell. Brinton, Laurel J. 2007. The development of I mean: Implications for the study of historical pragmatics. In: Fitzmaurice and Taavitsainen (eds.), 37–79. Brown, Roger, and Albert Gilman. 1960. The pronouns of power and solidarity. In: Thomas A. Sebeok (ed.), Style in Language, 253–276. Cambridge, MA: MIT Press. Brownlees, Nicholas. 1999. Corantos and Newsbooks: Language and Discourse in the First English Newspapers (1620–1641). Pisa: Edizioni ETS. Brownlees, Nicholas (ed.). 2006. News Discourse in Early Modern Britain. Selected Papers of CHINED 2004. Bern: Peter Lang. Burnley, David. 1983. A Guide to Chaucer’s Language. London: Macmillan. Busse, Beatrix. 2006. Vocative Constructions in the Language of Shakespeare. Amsterdam/ Philadelphia: John Benjamins. Busse, Ulrich. 2002. Linguistic Variation in the Shakespeare Corpus. Morpho-Syntactic Variability of Second Person Pronouns. Amsterdam/Philadelphia: John Benjamins. Busse, Ulrich. 2003. The co-occurrence of nominal and pronominal address forms in the Shakespeare corpus: Who says thou or you to whom? In: Taavitsainen and Jucker (eds.), 193–221. Craun, Edwin D. 1997. Lies, Slander, and Obscenity in Medieval English Literature. Pastoral Rhetoric and the Deviant Speaker. Cambridge: Cambridge University Press.
209
210
II Linguistic Levels Cruse, Alan. 2000. Meaning in Language. An Introduction to Semantics and Pragmatics. Oxford: Oxford University Press. Culpeper, Jonathan, and Elena Semino. 2000. Constructing witches and spells: Speech acts and activity types in Early Modern England. Journal of Historical Pragmatics 1(1): 97–116. Culpeper, Jonathan and Merja Kyto¨. 2010. Early Modern English Dialogues: Spoken Interaction as Writing. Cambridge: Cambridge University Press. Deutschmann, Mats. 2003. Apologising in British English. Umea˚: Institutionen fo¨r moderna spra˚k, Umea˚ University. Doty, Kathleen L. 2007. Telling tales: The role of scribes in constructing the discourse of the Salem witchcraft trials. Journal of Historical Pragmatics 8(1): 25–41. Doty, Kathleen and Risto Hiltunen. 2002. “I will tell, I will tell”: Confessional patterns in the Salem witch trials, 1692. Journal of Historical Pragmatics 3(2): 299–336. Fischer, Andreas. 1998. Marry. From religious invocation to discourse marker. In: Borgmeier, Grabes, and Jucker (eds.), 35–46. Fitzmaurice, Susan M. and Irma Taavitsainen (eds.). 2007. Methodological Issues in Historical Pragmatics. Berlin/New York: Mouton de Gruyter. Fritz, Gerd. 1995. Topics in the history of dialogue forms. In: Jucker (ed.), 469–498. Geis, Michael and Arnold M. Zwicky. 1971. On invited inferences. Linguistic Inquiry 2: 561–566. Gotti, Maurizio. 2006. Disseminating Early Modern science: Specialized news discourse in the Philosophical Transactions. In: Brownlees (ed.), 41–70. Herring, Susan C. 2003. Media and language change: Introduction. Journal of Historical Pragmatics (Special Issue on Media and Language Change, ed. by Susan C. Herring) 4(1): 1–17. Hiltunen, Risto. 2004. Salem, 1692: A case of courtroom discourse in a historical perspective. In: Risto Hiltunen and Shinichiro Watanabe (eds.), Approaches to Style and Discourse in English, 3–26. Osaka: Osaka University Press. Hiltunen, Risto. 2006. “Eala, geferan and gode wyrhtan”: On interjections in Old English. In: John Walmsley (ed.), Inside Old English: Essays in Honour of Bruce Mitchell, 91–116. Oxford: Balckwell. Hiltunen, Risto and Matti Peikola. 2007. Trial discourse and manuscript context: Scribal profiles in the Salem witchcraft records. Journal of Historical Pragmatics 8(1): 43–68. Honegger, Thomas. 2003. “And if ye wol nat so, my lady sweete, thanne preye I thee, […].”: Forms of address in Chaucer’s Knight’s Tale. In: Taavitsainen and Jucker (eds.), 61–84. Jacobsson, Mattias. 2002. Thank you and thanks in Early Modern English. ICAME Journal 26: 63–80. Jucker, Andreas H. 1997. The discourse marker well in the history of English. English Language and Linguistics 1(1): 91–110. Jucker, Andreas H. 2002. Discourse markers in Early Modern English. In: Richard Watts and Peter Trudgill (eds.), Alternative Histories of English, 210–230. London/New York: Routledge. Jucker, Andreas H. 2006. “Thou art so loothly and so oold also”: The use of ye and thou in Chaucer’s Canterbury Tales. Anglistik 17(2): 57–72. Jucker, Andreas H., Gerold Schneider, Irma Taavitsainen, and Barb Breustedt. 2008. Fishing for compliments: Precision and recall in corpus-linguistic compliment research. In: Jucker and Taavitsainen (eds.), 273–294. Jucker, Andreas H. and Irma Taavitsainen. 2000. Diachronic speech act analysis: Insults from flyting to flaming. Journal of Historical Pragmatics 1(1): 67–95. Jucker, Andreas H., and Irma Taavitsainen. 2008b. Apologies in the history of English: Routinized and lexicalized expressions of responsibility and regret. In: Jucker and Taavitsainen (eds.), 229–244. Jucker, Andreas H. (ed.). 1995. Historical Pragmatics. Pragmatic Developments in the History of English. Amsterdam/Philadelphia: John Benjamins. Jucker, Andreas H., Gerd Fritz, and Franz Lebsanft (eds.). 1999. Historical Dialogue Analysis. Amsterdam/Philadelphia: John Benjamins. Jucker, Andreas H. and Irma Taavitsainen (eds.). 2008a. Speech Acts in the History of English. Amsterdam/Philadelphia: John Benjamins.
13 Linguistic Levels: Pragmatics and discourse Kahlas-Tarkka, Leena and Matti Rissanen. 2007. The sullen and the talkative: Discourse strategies in the Salem examinations. Journal of Historical Pragmatics 8(1): 1–24. Keller, Rudi. 1994. On Language Change. The Invisible Hand in Language. London: Routledge. Kilian, Jo¨rg. 2005. Historische Dialoganalyse. Tu¨bingen: Niemeyer. Koch, Peter. 1999. Court records and cartoons: Reflections of spontaneous dialogue in Early Romance texts. In: Jucker, Fritz, and Lebsanft (eds.), 399–429. Kohnen, Thomas. 2000. Explicit performatives in Old English: A corpus-based study of directives. Journal of Historical Pragmatics 1(2): 301–321. Kohnen, Thomas. 2007. From Helsinki through the centuries: The design and development of English diachronic corpora. In: Pa¨ivi Pahta, Irma Taavitsainen, Terttu Nevalainen, and Jukka Tyrkko¨ (eds.), Studies in Variation, Contacts and Change in English, Vol. 2. http:// www.helsinki.fi/varieng/journal/volumes/02/ Kohnen, Thomas. 2008a. Directives in Old English: Beyond politeness? In: Jucker and Taavitsainen (eds.), 27–44. Kohnen, Thomas. 2008b. Tracing directives through text and time: Towards a methodology of a corpus-based diachronic speech-act analysis. In: Jucker and Taavitsainen (eds.), 295–310. Levinson, Stephen C. 1983. Pragmatics. Cambridge: Cambridge University Press. Ma¨kinen, Martti. 2004. Herbal recipes and recipes in herbals: Intertextuality in early English medical writing. In: Taavitsainen and Pahta (eds.), 144–173. Mazzon, Gabriella. 2000. Social relations and form of address in the Canterbury Tales. In: Dieter Kastovsky and Arthur Mettinger (eds.), The History of English in a Social Context. A Contribution to Historical Sociolinguistics, 135–168. Berlin/New York: Mouton de Gruyter. Mey, Jacob L. 1998. Pragmatics. In: Jacob L. Mey (ed.), Concise Encyclopedia of Pragmatics, 716– 737. Amsterdam: Elsevier. Nathan, N. 1959. Pronouns of address in the Canterbury Tales. Mediaeval Studies xxi: 193–201. Person, Raymond R., Jr. 2009. Oh in Shakespeare: A conversation analytic approach. Journal of Historical Pragmatics 10(1): 84–107. Raymond, Joad (ed.). 2006. News Networks in Seventeenth Century Britain and Europe. London: Routledge. Raymond, Joad. 2003. Pamphlets and Pamphleteering in Early Modern Britain. Cambridge: Cambridge University Press. Sommerville, John. 1996. The News Revolution in England. Cultural Dynamics of Daily Information. New York/Oxford: Oxford University Press. Sperber, Dan, and Ira A. Noveck. 2004. Introduction. In: Ira A. Noveck and Dan Sperber (eds.), Experimental Pragmatics, 1–22. Houndmills: Palgrave Macmillan. Stein, Dieter. 2003. Pronominal usage in Shakespeare: Between sociolinguistics and conversational analysis. In: Taavitsainen and Jucker (eds.), 251–307. Studer, Patrick. 2008. Historical Corpus Stylistics. Media, Technology and Change. London: Continuum. Taavitsainen, Irma. 1995. Interjections in Early Modern English: From imitation of spoken to conventions of written language. In: Jucker (ed.) 439–465. Taavitsainen, Irma. 1999. Dialogues in Late Medieval and Early Modern English medical writing. In: Jucker, Fritz, and Lebsanft (eds.), 243–268. Taavitsainen, Irma. 2006. Medical discourse: Early genres, 14th and 15th centuries. In: Keith Brown (ed.), Encyclopedia of Language and Linguistics, 2nd edn., 688–694. Oxford: Elsevier. Taavitsainen, Irma and Andreas H. Jucker (eds.). 2003. Diachronic Perspectives on Address Term Systems. Amsterdam/Philadelphia: John Benjamins. Taavitsainen, Irma and Andreas H. Jucker. 2007. Speech act verbs and speech acts in the history of English. In: Susan M. Fitzmaurice and Irma Taavitsainen (eds.), Methods in Historical Pragmatics, 107–138. Berlin: Mouton de Gruyter. Taavitsainen, Irma and Pa¨ivi Pahta (eds.). 2004. Medical and Scientific Writing in Late Medieval English. Cambridge: Cambridge University Press. Traugott, Elizabeth Closs. 2004. Historical pragmatics. In: Laurence R. Horn and Gregory Ward (eds.), The Handbook of Pragmatics, 538–561. Oxford: Blackwell.
211
212
II Linguistic Levels Traugott, Elizabeth Closs and Richard B. Dasher. 2005. Regularity in Semantic Change. Cambridge: Cambridge University Press. Trosborg, Anna. 1994. Interlanguage Pragmatics. Requests, Complaints and Apologies. Berlin/New York: Mouton de Gruyter. Ungerer, Friedrich (ed.). 2000. English Media Texts Past and Present. Amsterdam/Philadelphia: John Benjamins. Valkonen, Petteri. 2008. Showing a little promise: Identifying and retrieving explicit illocutionary acts from a corpus of written prose. In: Jucker and Taavitsainen (eds.), 247–272. Valle, Ellen. 1997. A scientific community and its texts: A historical discourse study. In: BrittLouise Gunnarsson, Per Linell, and Bengt Nordberg (eds.). The Construction of Professional Discourse, 76–98. London/New York: Longman. Valle, Ellen. 1999. A Collective Intelligence. The Life Sciences in the Royal Society as a Scientific Discourse Community, 1665–1965. Turku: University of Turku. Valle, Ellen. 2006. Reporting the doings of the curious: Authors and editors in the Philosophical Transactions of the Royal Society of London. In: Brownlees (ed.), 71–90. Verschueren, Jef. 1987. The pragmatic perspective. In: Marcella Bertuccelli-Papi and Jef Verschueren (eds.), The Pragmatic Perspective: Selected Papers from the 1985 International Pragmatics Conference, 3–8. Amsterdam/Philadelphia: John Benjamins. Verschueren, Jef. 1999. Understanding Pragmatics. London: Arnold. Watts, Richard J. 1999. Refugiate in a strange countrey: Learning English through dialogues in the 16th century. In: Jucker, Fritz, and Lebsanft (eds.), 215–241. Wilcockson, Colin. 1980. Thou and ye in Chaucer’s Clerk’s Tale. The Use of English 31(3): 37–43.
Andreas H. Jucker, Zu¨rich (Switzerland)
14 Linguistic Levels: Onomastics 1 2 3 4 5 6
Introduction Toponyms Anthroponyms Transmission of names Summary References
Abstract Names provide evidence for language history in two main respects: firstly, as regards lexical and semantic content when first coined; and secondly, as regards phonological and morphological development over the course of time. In neither respect is there widespread agreement as to the extent to which evidence from names can be extrapolated to other areas of language. On the one hand, both place-names and personal names testify to areas of vocabulary and registers of language sparsely represented in other sources; on the other, it is sometimes unclear whether these reflect ordinary language or a specialized onomastic usage. Factors pertaining to the formation and transmission of names are in some respects unique, and will be outlined in this chapter alongside a discussion of the main types of linguistic evidence preserved in the onomasticon. Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 212–223
14 Linguistic Levels: Onomastics
1 Introduction Onomastics is the study of names, its two main branches being toponymy (the study of place-names) and anthroponymy (the study of people’s names). Traditionally regarded as a sub-class of nouns having reference but no sense, names occupy a special position within language in that they can be used without understanding of semantic content. Partly for this reason, they tend to have a high survival rate, outlasting changes and developments in the lexicon, and easily being taken over by new groups of speakers in situations of language contact. Since most names originate as descriptive phrases, they preserve evidence for early lexis, often within areas of vocabulary sparsely represented in other sources. Many place-names, and some surnames, are still associated with their place of origin, so the data also contribute to the identification of dialectal isoglosses. Moreover, since names are generally coined in speech rather than in writing, they testify to a colloquial register of language as opposed to the more formal registers characteristic of documentary records and literary texts. Much research has been directed towards establishing the etymologies of names whose origins are no longer transparent, using a standard methodology whereby a comprehensive collection of early spellings is assembled for each name in order to trace its historical development. These spellings themselves can then be used to reveal morphological and phonological changes over the course of time, often illustrating trends in non-onomastic as well as onomastic language. The relationship between the two is not always straightforward, however, since the factors pertaining to the formation and transmission of names are in some respects unique. This chapter will discuss the main types of linguistic evidence preserved in names of various kinds, and will also consider the relevance of this evidence to other areas of language.
2 Toponyms 2.1 The origins of place-names The names of most villages, towns and cities in England were coined during the Anglo-Saxon period from Old English or (in areas of Scandinavian settlement) Old Norse. Others derive from the Celtic languages more strongly represented in Scotland, Wales and Ireland, while survivals from pre-Celtic linguistic strata are mainly found in the names of large topographical features. The names of major rivers are among the most ancient toponyms, and parallels between British and European river-names appear to reflect a system of hydronymy in use on the continent and brought to Britain by pre-Celtic immigrants. These river-names preserve evidence for the earliest form of language spoken in the British Isles, although it remains controversial whether this language was Indo-European – the majority view – or non-Indo-European. The names of smaller features tend to be later, dating from the medieval or early modern periods, and the same applies to the field-names given to units of cultivated land. Names for new urban developments continue to be created up to the present day, with street-names still being coined in large numbers. Nevertheless, even streetnames can be more than a thousand years old in the medieval parts of cities such as Derby, London, Nottingham and York.
213
214
II Linguistic Levels Research into the origins of English place-names has been carried out systematically on a county-by-county basis since the 1920s by the English Place-Name Society (1924–), and is published in a series of annual volumes known collectively as the English PlaceName Survey (EPNS). Whereas the Survey initially focused on names of historical significance such as medieval settlement-names, a growing recognition of the linguistic interest of the material led to coverage being expanded to include other types of names and those dating from later periods. This means that early volumes are not only less up-to-date than those currently being produced, but more limited in scope. Supplementary publications, including a field-name series, aim to redress the balance; and regional studies and dictionaries are also underway to provide coverage of the major names of all parts of England in advance of completion of the full Survey. Volumes 25–26 of the EPNS series comprise a dictionary of place-name terminology, a successor to which is currently in progress (Parsons, Styles, and Hough 1997–), and EPNS collections also form the basis for Field’s (1972) dictionary of field-names.
2.2 The structure of place-names Most English settlement-names are made up of one or (more usually) two elements, identical or closely related to vocabulary words. Most represent a description of a landscape feature or man-made structure, as with the single element or “simplex” names Dean or Deane (OE denu ‘main valley’), Ford (OE ford ‘ford, river-crossing’), Ham (OE hamm ‘hemmed-in land’), Hope (OE hop ‘small enclosed valley’), Lea or Leigh (OE le¯ah ‘wood, clearing’), Stoke (OE stoc ‘outlying farmstead’), Wick (OE wı¯c ‘specialized farm’) and Worth (OE worth ‘enclosure’). The corpus is highly repetitive, and each of the above examples occurs several times in different parts of England. Compound names provide a more precise description by including information on such aspects as appearance, ownership, usage, flora, or fauna. Examples include Abbotsley ‘Ealdbeald’s clearing’, Bagley ‘badger clearing’, Bradford ‘broad ford’, Bulwick ‘bull farm’ and Cotterstock ‘dairy farm’. Here, as in most non-Celtic place-names, the descriptive element or “specific” precedes the defining element or “generic”. Whereas the generic usually identifies a topographical feature or habitation, specifics have a much wider range, including personal names, older place-names, descriptive adjectives, animal, bird, and plant names, and occupational terms. Some names also contain an additional element generally known as an “affix” (although the term “distinguisher” might be more accurate), as with Stoke Mandeville, from the Mandeville family, and Stoke on Trent, from its position on the River Trent. In field-names and street-names, the generic is often a term for a field or street, with specifics again being more varied. In London, Chancery Lane refers to the chancellor’s office (ME chauncerie), Mincing Lane to nuns (OE myncen), and Sherborne Lane to a privy (ME *shite-burgh), while Mansfield Hill and Markfield Road preserve earlier field-names meaning ‘common field’ (OE (ge)mæ¯ne) and ‘boundary field’ (OE mearc) respectively (Mills 2001). However, non-literal formations are also common. Bare Arse, Labour in Vain, and Small Gains are recurrent field-names referring to unproductive land; irony may be suspected in some of the many fields called Paradise; the ubiquitous Hundred Acres almost invariably designates a very small field; and at least one Mount Pleasant refers to a refuse tip. A tendency for landscape features to receive metaphorical names is illustrated by the recurrent Cow and Calf Rocks (large and
14 Linguistic Levels: Onomastics small), and by various hill-names from the Scandinavian-derived carline ‘old woman’, including Carling Howe, Carling Knott and Fishcarling Head.
2.3 The language of place-names Many place-names are coined from terms also on record elsewhere, while others contain elements that are otherwise unknown, and for which the only evidence is the toponymicon. The repetitive nature of the corpus makes it possible to assemble a collection of names containing the same term in order to analyze its range of use. This applies particularly to generics. The most common element in English settlement-names, OE tu¯n ‘farmstead, village’, occurs in hundreds of names, and some of those mentioned above in more than a hundred. Many topographical generics can still be compared directly with the features described, and this has led to one of the main insights of place-name scholarship in recent years. The examination of places named from the same generic has identified subtle distinctions between the use of terms previously thought to be synonyms, revealing that the Anglo-Saxons had a more extensive and nuanced vocabulary for landscape features than has survived into later stages of English (Gelling and Cole 2000). Definitions given in Section 2.2 are summary only, and do not fully reflect the distinction between, for instance, OE denu, the standard term for a main valley, and OE hop, a small and often remote enclosed valley. Both differ from OE cumb ‘short, broad valley with three fairly steep sides’, OE slæd ‘flat-bottomed valley’, OE halh ‘nook’ (often referring to a less firmly-shaped valley than cumb or denu), and so on. Similarly precise usages have been established for other areas of topographical lexis, including terms unattested outside the place-name corpus such as OE *hlenc ‘extensive hill-slope’ and OE *ofer ‘flat-topped ridge with a convex shoulder’. Evidence from other Germanic languages may assist interpretation. A meaning of OE halh as ‘slightly raised ground in marsh’ is suggested by some place-name occurrences and supported by a similar use of the North Frisian cognate. Specifics represent a wider cross-section of open class vocabulary than generics, and contain a higher proportion of unattested words. Again the profile of use throws light on meaning. The specific of Bagley is an unattested OE *bagga. The range of generics with which it combines, including references to natural habitat and snares, as in Bag Hill (OE hyll ‘hill’), Bagshot (OE sce¯at ‘projecting land’) and Bawdrip (OE træppe ‘trap’), suggests a wild animal such as the badger, and this is supported by Germanic cognates. The specific of Grazeley (OE sol ‘wallowing-place’), Gresty (OE stı¯g ‘path’) and Greywell (OE wella ‘spring, stream’) is a substantive use of OE græ¯g ‘grey (animal)’ thought to refer to the wolf, as does an OE *wearg in place-names such as Warnborough and Wreighburn (both OE burna ‘stream’). Other words are attested in the place-name corpus earlier than in literary sources. OE *bagga may be the root of PDE badger, although this is uncertain. The specific of Bulwick is an OE *bula, the etymon of PDE bull, on independent record only from c.1200. Carling Howe is first recorded c.1170, whereas the earliest occurrence of carline ‘old woman’ recorded in the OED dates from c.1300. The compound *fish-carline ‘fish-wife’, attested in Fishcarling Head, is otherwise unknown. In some instances, place-names support other types of evidence for word meanings. The Dictionary of Old English (Cameron et al. 2008) tentatively suggests a meaning “ ‘churn, or even ‘dairy’ ” for three gloss occurrences of OE corþer, and this is confirmed
215
216
II Linguistic Levels by a toponymic use as the specific of Cotterstock (OE stoc ‘farm’). A substantive use of OE bru¯n ‘brown’ recorded as a nonce occurrence within a riddle appears to refer to a brown animal, possibly the pig, and again this is confirmed by its use in place-names such as Broomden (OE denn ‘woodland pasture, especially for swine’) and Brownwich (OE wı¯c ‘specialized farm’). Certain types of words, including animal names, bird names and topographical terms, are particularly well represented in place-names. So too are plant names (Hough 2003). The toponymic corpus considerably extends our knowledge of these areas of lexis, as of the colloquial range of vocabulary. Several of the above examples may represent demotic terms as opposed to the more formal registers of language represented in written sources. However, it is unclear to what extent the place-name corpus as a whole is representative of ordinary vocabulary. Some elements, particularly generics, are thought to belong to an onomastic register common to the northwest Germanic languages, which may have diverged from ordinary lexis at an early date (Nicolaisen 1995). Parallels between OE halh and its North Frisian cognate are mentioned above, and there are similar instances where pairs of generics in the West and North Germanic toponymica have more in common with each other than with the corresponding lexical terms. OE ha¯m ‘homestead, estate’, one of the earliest habitative generics in English place-names, is cognate with ON heimr, and this may help to explain why it appears to have been in use as a place-name forming element earlier than as a common noun. Other generics also have a different chronological or geographical profile from their lexical counterparts, while the fact that some elements (e.g. OE le¯ah) occur as generics only, apparently not being available for use as specifics, further supports the theory that they were not selected freely from the lexicon. It may therefore be appropriate to regard such elements as cognate with, rather than identical to, vocabulary items, deriving from a common ancestor but developing along different lines. When they began to develop separately is uncertain, but evidence from river-names suggests that a so-called “onomastic dialect” may already have emerged early in the history of Indo-European (Kitson 1996). Specifics may be closer to ordinary lexis, but again develop uses in place-names which are not necessarily the same as those in non-onomastic language. Analogy is common in place-name formation, and this limits the value of toponymic evidence for word geography as well as for historical semantics. The place-name distribution of carline, for instance, suggests a wider currency in northern dialects of Middle English than is supported by lexical evidence, extending to the north-west of England outside the Danelaw, but it is possible that this is due to the conventional use of the term for landscape features. Another factor is the high level of repetition within the name stock, which suggests that even some compounds may have been drawn from an existing pool. Ekwall (1960) includes 40 occurrences of Burton from OE burh-tu¯n ‘settlement by a fortification’, 29 of Charlton from OE ce¯orla-tu¯n ‘settlement of freemen’, and 20 of Easton from OE e¯ast-tu¯n ‘east settlement’, and it is unlikely that they were coined afresh on each occasion. Field-names and street-names preserve much vocabulary from the medieval and later periods, and comprise the bulk of the evidence for Middle English terms attested uniquely or earliest in place-names (Hough 2002). Street-names are a particularly rich source of occupational terms. Birchin Lane in London, recorded from the 12th century,
14 Linguistic Levels: Onomastics derives from an unattested ME *berdcherver ‘beard-cutter, barber’; Felter Lane in York, recorded from the 13th century, contains the term felter ‘felt-maker’ attested in the OED only from 1605; and Fletcher Gate in Nottingham, recorded in 1335 as Flesshewergate, is the earliest occurrence cited by the OED of the obsolete term flesh-hewer ‘butcher’. Again there is much repetition within the corpus. Some names are transferred directly from one place to another (e.g. Piccadilly, now found in Manchester and York as well as London), while some are coined by analogy with others. Street-names in -gate are common in areas of urban settlement within the former Danelaw, where they often derive from ON gata ‘street’; but in other instances they have simply been modelled on the original Viking names (Fellows-Jensen 2007). They cannot therefore be used as evidence for the currency of the term gata itself, nor for Scandinavian influence on the language. Nonetheless, microtoponyms make an important contribution to the study of Old and Middle English dialectology by offering a large quantity of data for analysis. On a national scale, they facilitate the mapping of synonyms, an approach taken in Kitson’s (1995) investigation of Old English dialect isoglosses, and in studies of the distribution of complementary terms such as whin, gorse, and furze (e.g. Cameron 2008). On a regional scale, they reflect a mix of languages more accurately than the smaller number of settlement names. This has been exploited in studies using the ratio between fieldname elements of Old English and Old Norse origin to measure the extent of Scandinavian impact on local dialect (e.g. Cameron 1996, Watts 2002, Parsons 2006).
3 Anthroponyms 3.1 Personal names Like place-names, Anglo-Saxon personal names are made up of one or (more usually) two elements drawn from a corpus corresponding closely to a subset of the lexicon. Animal names, abstract nouns, and descriptive adjectives are particularly well represented, as is the semantic field of warfare. Masculine and feminine names are formed along parallel lines, with the grammatical gender of the second element or “deuterotheme” often coinciding with the gender of the person. Examples include Æthelgar ‘noble-spear’, Guthfrith ‘battle-peace’, and Wulfstan ‘wolf-stone’ (all masculine), Æthelflæd ‘noblebeauty’, Æthelthryth ‘noble-strength’, and Wulfgifu ‘wolf-gift’ (all feminine). These and other names do not make literal sense, but may have signalled family relationships. Members of a kin group characteristically have names with alliterating (sometimes identical) first elements or “protothemes”, while the deuterotheme of a parent’s name may be passed on to children of the same sex. The naming system goes back to Common Germanic. Links have been identified with the vocabulary of heroic poetry, and here even more than with place-names there is reason to regard name elements as cognate with, rather than identical to, their lexical counterparts. Whereas most place-names were literally descriptive of their referents in Anglo-Saxon England, personal names clearly were not, and this lack of sense suggests that elements were drawn from an anthroponymicon rather than from the lexicon. The impression of an older system reflected but no longer motivated in the Anglo-Saxon naming tradition is strengthened by instances where the correlation between natural and grammatical gender is not consistently maintained.
217
218
II Linguistic Levels Elements corresponding to feminine nouns such as OE mund ‘protection’ and OE no¯th ‘boldness’ form masculine names (e.g. Byrhtnoth ‘bright-boldness’, Sigemund ‘victoryprotection’), as do elements corresponding to neuter nouns such as OE cild ‘young person’ (e.g. Leofcild ‘dear-child’) and to adjectives such as OE beald ‘bold’ and OE beorht ‘bright’ (e.g. Æthelbeald ‘noble-bold’, Ealdbeorht ‘old-bright’). Although some elements function both as protothemes and as deuterothemes, others appear to be restricted to one position only, and this lack of flexibility – together with the fact that not all vocabulary words appear to have been available for use in coining names – again points to some fossilization of the system. This means that where a name element is unattested in literary sources from the Anglo-Saxon period, as with *flæd, whose meaning is reconstructed from Germanic cognates as “elegance or daintiness as of a courtly lady” (Kitson 2002: 97), it cannot be assumed that it was ever in use as a lexical item in Old English. Anglo-Saxon personal names were largely replaced by Continental Germanic and Biblical names after the Norman Conquest, and are sparsely represented in the present-day name stock. Although personal names have occasionally been coined from vocabulary words, as with the “virtue” names associated with the Puritan movement in post-Reformation England (e.g. Hope, Joy, Patience), and others referring to precious stones, plants, and the like (e.g. Ruby, Heather, Holly), the vast majority of names in current use are of Biblical or classical origin, and even those that remain semantically transparent are generally used without reference to etymological meaning. Recent research has emphasized the importance of social and cultural factors in present-day and early modern name giving, with linguistic origin being a lesser consideration.
3.2 Bynames Prior to the evolution of surnames, bynames were used to differentiate between people with the same personal name. Already in use in Anglo-Saxon England, they became much more common during the early Middle English period, partly because of the dwindling number of baptismal names in general use. By the late 13th century, most personal names were routinely qualified by a byname. The four main categories of bynames are local, familial, occupational, and characteristic. Examples from Clark’s (1983) study of King’s Lynn bynames are atte Ling ‘beside the heather’ (local), Lellesmai ‘Lelle’s kinsman’ (familial), Le Blekestere ‘the bleacher’ (occupational) and Le Longe ‘the tall’ (characteristic). Bynames differ from the other types of names discussed above in being literally descriptive, and drawn freely from the lexicon. Despite having much in common with surnames, they were not hereditary (nor shared with siblings), and hence did not lose their lexical meaning. Because of this, they can be taken to represent contemporary language rather than fossilized forms. Nonetheless, they may be difficult to interpret. Bynames such as Ioie ‘joy’ and Trouthe ‘truth’ may be understood literally or ironically, and Clark discusses ambiguities surrounding others such as Baril ‘barrel’ and Peper ‘pepper’, which may be occupational referring to a barrel-maker and spicer, or characteristic referring to corpulence and hot temper. Like place-names, bynames preserve a number of words unattested elsewhere or on record only from a later date. Tengvik’s seminal study of Old English bynames identifies 58 antedated words (Tengvik 1938: 24–26), and others have since been added. Some are animal- or bird-names used as characteristic bynames. Otherwise unknown terms
14 Linguistic Levels: Onomastics include OE *gor-pyttel ‘dung-hawk’, recorded as the byname of Gotselin Gorpittel (c.1100–1130). Occupational bynames are an important source of Middle English occupational terms, and several of the King’s Lynn bynames either antedate or are close to antedating the earliest documentation in literary sources. These include Oylman (a worker at the town’s oil-mills) and Habertasker ‘haberdasher’. Previously unrecorded occupational terms from the same corpus include *candelwif and *chesewoman, referring to women involved in the candle- and cheese-making industries. Topographical terms are preserved in local bynames, again often providing antedatings (e.g. Carlsson 1989: 146–147). The languages represented in bynames also make it possible to trace linguistic influences on local dialect. Le Blekestere is based on a Scandinavian loanword, while trading links with France are reflected in the definite article and in French forms such as Baril.
3.3 Surnames Like bynames, surnames evolved to differentiate between people with the same personal name; but unlike bynames, they were passed on from one generation to the next, functioning as markers of relationship rather than as descriptions of individuals. Surnames began to come into use in England during the late 11th century, partly as a result of changes in society following the Norman Conquest, but the system was not fully established until about 1400. The main types of surnames correspond to the main types of bynames: local, familial, occupational, and characteristic. The earliest were local surnames held by landowning families, some imported from Normandy and others taken from the names of estates in England. The use of surnames gradually spread down the social scale, with occupational surnames associated particularly with skilled craftsmen, and characteristic surnames with the lower classes. Local surnames – those referring to place of origin, or to ownership of land – can themselves be divided into two categories: locative and topographical. Locative surnames derive from place-names, topographical surnames from lexical words. Thus the surname Glasgow is locative, from the Scottish city, while the surname Hough is topographical, from OE ho¯h ‘spur of land’. The latter type can reflect regional variation in morphology as well as lexis. Examples such as Bridger and Bridgeman ‘dweller by the bridge’ and Weller and Wellman ‘dweller by the stream’ show the morphemes -er and -man suffixed to a topographical term to indicate place of residence. This type of surname is common in central and southern England, but rare in the north, where the morphemes do not seem to have been productive in this sense. Familial surnames, or surnames of relationship, most commonly derive from parents’ forenames. Those from a father’s name are known as “patronymics”; those from a mother’s name as “metronymics”. The earliest derive from unaffixed personal names, whether from Old English, as with Godwin, or from the continental name stock introduced to England after the Conquest, as with Allen, Maude, and Thomas. Surnames with the suffix -son are slightly later and often formed from shortened, or hypochoristic, forms of personal names, as with Dobson and Robson (Robert), Ibbotson and Ibson (Isabel), and Wilson and Wilkinson (William). Again there are regional variations. Most surnames in -son originated in the north and north midlands of England, and in southern Scotland. In southern and central England, the suffix -s was more common, as in Andrews, Roberts and Williams.
219
220
II Linguistic Levels Occupational surnames reflect lexical variation in different areas (McKinley 1990: 143–147). Surnames relating to the same trade within the cloth industry are Fuller (south and south-east), Tucker (south-west), and Walker (north); while Barker (north) and Tanner (south) are synonyms. This group is also a primary source of information on Middle English occupational terms. Thuresson (1950) deals with about 850 words, of which 271 were not recorded in the OED, and Fransson (1935) also provides a number of new words and predatings. Some surnames from characteristics are based on English or French phrases which appear to have had a wide currency in Middle English despite being rarely if at all recorded elsewhere. Examples such as Fairwether, Goodall ‘good ale’, Makehate ‘make joy’, Parlebien ‘well spoken’, Passavaunt ‘go before’, Proudfoot, and Spendlove are suggested by McKinley (1990: 166) to “perhaps convey some flavour of the ordinary man’s spoken English as it was in the 13th and 14th centuries, when most of these names arose”.
4 Transmission of names The lack of sense that is widely considered a defining characteristic of names allows them to preserve fossilized forms of words that have long since disappeared from the lexicon. Moreover, the fact that most place-names are locatable in space, and many personal names locatable in time, gives them an advantage over texts preserved in manuscripts of uncertain provenance. Forms of moneyers’ names on Anglo-Saxon coins can be ordered chronologically by coin-types, providing closely datable material for studies of Old English phonology (e.g. Colman 1984, 1992), while the range of spellings relating to individual place-names and place-name elements can reveal dialectal and morphological variation. In East Anglia, spellings in place-names from OE stre¯t/stræ¯t ‘Roman road’ (e.g. Stradsett, Stradbroke, Stradishall) testify to a development from Saxon æ¯-forms rather than Anglian e¯-forms in early Old English (Kristensson 2001); variation between Scandinavian and English lexical terms in spellings of individual topographical names from the 14th to 16th centuries throws light on local dialect during the Middle English period (Sandred 2001: 51); and 13th- and 14th-century forms of the field-names Hanging Furlong and Hanging Wong with the present participle in reflect Scandinavian morphological influence (Sandred 2001: 51–52). This type of evidence needs to be handled with care, however, as it is also widely recognized that names may behave differently from vocabulary items. Lass (1973) points out that many surnames could not have developed their modern form through standard processes of phonological change, and illustrates this through a detailed analysis of the surname Shuttlebotham and its variants Shipperbottom, Shuflebotam, Shovelbottom, Shoebotham, Shoebottom, and Shubotham. Colman (1988) too urges caution in the use of names as evidence for linguistic reconstruction and historical dialectology. Some name spellings become fossilized, failing to reflect phonological variation, while others reflect the influence of more than one etymon. Here as in the initial formation of names, analogy plays a key role. Following the loss of semantic meaning, element substitution may occur, with familiar elements tending to replace less familiar ones. The common element wulf ‘wolf’, often spelled in Old English personal names, replaces the original deuterotheme col ‘coal’ of the moneyers’ names Sæcol
14 Linguistic Levels: Onomastics and Swartcol in the later spellings and ; the topographical placename generic OE le¯ah, often surviving as , influences the development of names such as Hawksley and Notley from hafoces hlewe ‘hawk’s tumulus’ and hnut clyf ‘nut slope’ (Gelling 1997: 202–205), and of Grazeley from grægsole; and the development of Warnborough from weargeburnan shows interference from the habitative generic OE burh ‘fortification’. It is particularly common for place-names to be influenced by neighbouring ones, and Coates (1987: 329) cites 24 pairs of place-names, mostly less than five miles apart, where one has developed along non-standard lines through the influence of the other. Analogy also operates at the level of the whole name. The ubiquitous Burton, from OE burh-tu¯n, has attracted to itself place-names from different specifics. Burton Bradstock contains the name of the River Bride, Burton in Sussex contains an Old English personal name Budeca, and Burton Salmon contains an adjective OE bra¯d ‘broad’. Similarly, a Middlesex place-name containing a personal name Ceolred has developed into another Charlton, and a Devon place-name containing a personal name Ælfric or Æthelric into another Easton. In each case, the influence of analogy has overridden standard processes of phonological change. Spelling patterns may also be affected by the medium in which the names are recorded, an area explored by Anderson and Colman (2004) in relation to wrap-around conventions and the transposition of graphs in the representation of personal names on Anglo-Saxon coins. Another factor is the role of government in regulating coinage and documenting names. The recurrence of certain forms on coins may reflect not pronunciation but an attempt to regularize spelling (Smart 1983); while students of placenames discriminate carefully between local spellings and those emanating from a centralized administration. The compound structure characteristic of both personal and place-name formations may also lead them to develop differently from their lexical cognates. In some instances, phonological changes are reflected earlier than in written texts. The loss of /d/ from word-final /nd/, common in written texts from the 15th century onwards, is attested as early as the late 12th and 13th centuries in the stressed (first) syllable of placenames and surnames, where it is attributed to the influence of the consonant initial in the second part of the compound (Wełna 2005). Consonant loss at the juncture of compound personal names is discussed alongside other issues specific to compounds by Colman (1984: 126–136), who demonstrates the value of Old English name-spellings as evidence for phonological developments in obscured compounds (i.e. compounds that have undergone phonological reduction and semantic obscuration). In general, motivated change affects names more than ordinary lexis. Personal names and surnames are subject to idiosyncratic choices as regards both pronunciation and spelling, and place-names too may be “improved”, as with Sherborne Lane in London mentioned above. Early spellings show that it preserves an unattested ME *shite-burgh ‘privy’, but this has been changed for reasons of delicacy. In other instances, an element that has become semantically opaque is reshaped to conform with known lexis, or simply replaced by a word that appears to make more sense. Mincing Lane in London derives from the obsolete word myncen ‘nun’, and Fletcher Gate in Nottingham from flesh-hewer ‘butcher’ – apparently associated with the later occupational term fletcher ‘arrow-maker’, itself now obsolete but plausible as a street-name specific. The same process of folk etymology accounts for the development (from a
221
222
II Linguistic Levels Lancashire place-name) of the surnames Shovelbottom, Shoebottom, and so on. So too the loss of much of the Anglo-Saxon personal name stock following the Norman Conquest led to the names becoming unfamiliar and being reshaped where they appear in surnames or place-names. The surname Freelove derives from an Old English personal name Frithulaf, while the first element of Abbotsley is an Old English personal name Ealdbald. Because names are so often affected by folk etymology, they do not always follow standard patterns of phonological or morphological development. For the same reason, however, they provide evidence for the operation of folk etymology itself.
5 Summary In sum, the relationship between names and lexis is not straightforward, but this does not diminish the value of onomastic material in the study of historical linguistics. Although neither the initial formation nor the subsequent transmission of names directly parallels the lexicon, the differences are themselves enlightening and reveal information unavailable from other sources. Handled with appropriate caution, onomastic evidence provides insights which both supplement and extend those offered by other areas of language.
6 References Anderson, John and Fran Colman. 2004. Non-rectilinear name-forms in Old English and the media of language. In: Gunnar Bergh, Jennifer Herriman, and Mats Moba¨rg (eds.), An International Master of Syntax and Semantics: Papers Presented to Aimo Seppa¨nen on the Occasion of his 75th Birthday, 31–42. Go¨teborg: Acta Universitatis Gothoburgensis. Cameron, Angus, Ashley Crandall Amos, Antonette diPaolo Healey et al. 2008. Dictionary of Old English: A-G on CD-ROM. Fascicle G and Fascicles A to F (with revisions). Toronto: Dictionary of Old English Project. http://www.doe.utoronto.ca/pages/pub/fasc-a-g.html Cameron, Jean. 2008. The distribution of whin, gorse and furze in English place-names. In: O. J. Padel and David N. Parsons (eds.), A Commodity of Good Names: Essays in Honour of Margaret Gelling, 253–258. Donington: Shaun Tyas. Cameron, Kenneth. 1996. The Scandinavian element in minor names and field-names in northeast Lincolnshire. Nomina 19: 5–27. Carlsson, Stig. 1989. Studies on Middle English Local Bynames in East Anglia. Lund: Lund University Press. Clark, Cecily. 1983. The early personal names of King’s Lynn: an essay in socio-cultural history. Part II – by-names. Nomina 7: 65–89. Coates, Richard. 1987. Pragmatic sources of analogical reformation. Journal of Linguistics 23: 319–340. Colman, Fran. 1984. Anglo-Saxon pennies and Old English phonology. Folia Linguistica Historica 6: 91–143. Colman, Fran. 1988. What is in a name? In: Jacek Fisiak (ed.), Historical Dialectology: Regional and Social, 111–137. Berlin/New York: Mouton de Gruyter. Colman, Fran. 1992. Money Talks. Berlin/New York: Mouton de Gruyter. Ekwall, Eilert. 1960. The Concise Oxford Dictionary of English Place-Names. 4th edn. Oxford: Clarendon Press. English Place-Name Society. 1924–. The Survey of English Place-Names. Vols. 1–. Cambridge/ Nottingham: Cambridge University Press and English Place-Name Society. Fellows-Jensen, Gillian. 2007. The Scandinavian element gata outside the urbanised settlements of the Danelaw. In: Beverley Ballin Smith, Simon Taylor, and Gareth Williams (eds.), West Over
14 Linguistic Levels: Onomastics
223
Sea: Studies in Scandinavian Sea-Borne Expansion and Settlement Before 1300. A Festschrift in Honour of Dr Barbara E. Crawford, 445–459. Leiden/Boston: Brill. Field, John. 1972. English Field-Names. A Dictionary. Newton Abbot: David & Charles. Fisiak, Jacek and Peter Trudgill (eds.). 2001. East Anglian English. Cambridge: D. S. Brewer. Fransson, Gustav. 1935. Middle English Surnames of Occupation 1100–1350. Lund: Gleerup. Gelling, Margaret. 1997. Signposts to the Past: Place-Names and the History of England. 3rd edn. Chichester: Phillimore. Gelling, Margaret and Ann Cole. 2000. The Landscape of Place-Names. Stamford: Shaun Tyas. Hough, Carole. 2002. Onomastic evidence for Middle English vocabulary. In: Peter J. Lucas and Angela M. Lucas (eds.), Middle English from Tongue to Text. Selected Papers from the Third International Conference on Middle English: Language and Text, held at Dublin, Ireland, 1–4 July 1999, 155–167. Frankfurt: Peter Lang. Hough, Carole. 2003. Place-name evidence for Anglo-Saxon plant-names. In: C. P. Biggam (ed.), From Earth to Art. The Many Aspects of the Plant-World in Anglo-Saxon England, 41–78. Amsterdam/New York: Rodopi. Kitson, Peter R. 1995. The nature of Old English dialect distributions, mainly as exhibited in charter boundaries. In: Jacek Fisiak (ed.), Medieval Dialectology, 43–135. Berlin/New York: Mouton de Gruyter. Kitson, Peter R. 1996. British and European river-names. Transactions of the Philological Society 94: 73–118. Kitson, Peter R. 2002. How Anglo-Saxon personal names work. Nomina 25: 91–131. Kristensson, Gillis. 2001. Language in contact: Old East Saxon and East Anglian. In: Fisiak and Trudgill (eds.), 63–70. Lass, R. 1973. Review of P. H. Reaney, The Origin of English Surnames, Routledge and Kegan Paul, London, 1967. Foundations of Language 9: 392–402. McKinley, Richard. 1990. A History of British Surnames. London/New York: Longman. Mills, A. D. 2001. A Dictionary of London Place Names. Oxford: Oxford University Press. Nicolaisen, W. F. H. 1995. Is there a Northwest Germanic toponymy? Some thoughts and a proposal. In: Edith Marold and Christiane Zimmermann (eds.), Nordwestgermanisch, 102–114. Berlin/New York: Walter de Gruyter. Parsons, David N. 2006. Field-name statistics, Norfolk and the Danelaw. In: Peder Gammeltoft and Bent Jørgensen (eds.), Names Through the Looking-Glass: Festschrift in Honour of Gillian Fellows-Jensen July 5th 2006, 165–188. Copenhagen: C. A. Reitzels. Parsons, David N. and (for Vols. 1 and 2) Tania Styles with (for Vol. 1) Carole Hough. 1997–. The ´ –Box); 2 (Brace–Cæster); 3 (Ceafor–Cock-pit). NotVocabulary of English Place-Names: 1 (A tingham: Centre for English Name Studies (Vols. 1 and 2) and English Place-Name Society (Vol. 3). Sandred, Karl Inge. 2001. East Anglian place-names: Sources of lost dialect. In: Fisiak and Trudgill (eds.), 39–61. Smart, Veronica. 1983. Variation between Æthel- and Ægel- as a name-element on coins. Nomina 7: 91–96. Tengvik, Go¨sta. 1938. Old English Bynames. Uppsala: Almqvist & Wiksells. Thuresson, Bertil. 1950. Middle English Occupational Terms. Lund: Gleerup. Watts, Victor. 2002. Medieval field-names in two South Durham townships. Nomina 25: 53–64. Wełna, Jerzy. 2005. “Now you see it, now you don’t” once more: The loss and insertion of dental stops in medieval English. Studia Anglica Posnaniensia 41: 71–84.
Carole Hough, Glasgow (UK)
224
II Linguistic Levels
15 Linguistic Levels: Orthography 1 Theoretical approaches to language, speech, and writing: between referentiality and autonomy 2 Definitions of orthography and related terms 3 Classification of writing systems and principles governing English orthography 4 Units of writing systems: terminological evolution 5 The inventory and distribution of English graphemes from the synchronic and diachronic perspective 6 Orthography as the source of phonological evidence 7 Sociolinguistic aspects of orthography 8 Summary 9 References
Abstract This chapter offers a critical overview of some of the most influential ideas concerning writing in general, and orthographic systems, with particular attention paid to English orthography and its intricate structure. It also presents and explains the terminology which can be found in literature dealing with this subject. Section 1 gives an account of different attitudes to writing and the two main theoretical approaches, relational and autonomistic, which they have motivated. It is followed by a summary of selected definitions of orthography and related terms in Section 2. Section 3 examines various types of writing systems, places the English orthographic system in that taxonomic context, and discusses its governing principles. Section 4 deals with the evolution of terms used to denote units of writing systems. In the Section 5 all the graphemes of English orthography are listed and selected historical aspects of their evolution are mentioned. Section 6 contains a brief evaluation of orthography as the source of evidence for phonological change. Finally, the sociolinguistic aspects of orthography are identified, including its role as a binding norm within a language community, and the significance of orthographic variation.
1 Theoretical approaches to language, speech, and writing: between referentiality and autonomy In order to provide a definition of orthography and put it in the appropriate theoretical context, one must first consider and define the general terms to which it is related. These are, most importantly, “language”, “speech”, and “writing”. It seems quite clear at present that speech and writing differ in numerous ways, including, for instance, the purely physical properties, as well as the communicative situations and the purposes for which each of them is used. Thus, speech is time-bound, fleeting, spontaneous, used to express opinions and emotions, whereas writing is space-bound, permanent, planned, edited, and used to record and convey information (detailed accounts of these differences can be found, e.g., in Crystal 1995: 291–293 and Cook 2004: 31–53). In fact, both speech and writing can be treated as complementary ways of using language, where the latter is understood as “a complex system residing in our brain which allows Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 224–237
15 Linguistic Levels: Orthography us to produce and interpret utterances” (Rogers 2005: 2). However, there is no general agreement among linguists as regards the relation between speech and language on one hand, and writing and language, as well as writing and orthography, on the other. Since the beginning of the 20th century, these notions have been viewed from two main points of view labelled as “relational” and “autonomistic” by Sgall (1987: 2–3; see also Ruszkiewicz 1976: 37–44). The structuralists, representing the relational perspective, equated speech with language, and considered writing as an extra-linguistic phenomenon: – “Language and writing are two distinct systems of signs; the second exists for the sole purpose of representing the first” (Saussure 1993: 41a). – “The written forms are secondary symbols of the spoken ones – symbols of symbols” (Sapir 1921: 20). – “Writing is not language, but merely a way of recording language by means of visible marks” (Bloomfield 1933: 21). – “The linguist distinguishes between language and writing” (Hockett 1958: 4). These quotations, to some extent, echo the classical theories of writing, which assumed a close dependence of written language on spoken language. For example, “[i]n De Interpretatione, Aristotle (1963, 43) states that ‘spoken sounds are symbols of affections of the soul, and written marks [are] symbols of spoken sounds’ ” (Liuzza 1996: 35). Also Quintilian claimed that “spelling should follow pronunciation because the text is a repository for the vox” (1920, 1:144; quoted in Liuzza 1996: 35). Because of the influence exerted by Saussure, Bloomfield, and other early structuralists, the validity of their statements has mostly been taken for granted. According to Venezky (1970: 27), “Bloomfield is responsible probably more than any other contemporary linguist for the view that writing is secondary and subservient to speech”, because he consistently expressed this opinion in several publications. In contrast to Saussure, Bloomfield, and others, scholars adhering to the autonomistic point of view, e.g., Vachek, Bolinger, Stetson, McIntosh, McLaughlin, and Venezky, have claimed that “[m]uch is written that is not pronounced” (Stetson 1981 [1937]: 35), and that “writing does more than represent speech” (Cook 2004: 32). More precisely, according to them “[w]riting is any manifestation of language in visible signs; a written language is a code that may not need preliminary decipherment into speech to be understood.” (Vachek 1982: 38). In this view, the written language “has to a great extent become an instrument for the direct expression of meaning, co-ordinate with audible language. The result of this has been that the written language has in part been developed on lines of its own, independent of the development of oral speech.” (Bradley 1919 [1913]: 14–15; quoted by Liuzza 1996: 27). Although the preoccupation with spelling as the reflection of speech was the dominant feature of discussions on orthography in the 20th century, the adherents of the autonomistic approach to writing also had some forerunners to draw upon. Already in the 16th and 17th centuries, some scholars (e.g. William Bullokar, Alexander Gil, Alexander Hume, and John Wallis) realized that orthography carried some non-phonemic information. Venezky (1970: 18, 23–25) mentions also the writings of Goold Brown (especially his comprehensive grammar published in 1850), which clearly expressed the idea that “words are not mere sounds, and in their orthography more is implied than in phonetics
225
226
II Linguistic Levels or phonography. Ideographic forms have, in general, the advantage of preserving the identity, history, and lineage of words” (quoted in Venezky 1970: 18, 24). Brown did not deny the close relationship between letters and sounds, but he also emphasized the relative autonomy of the writing system: “The deaf and the dumb, also, to whom none of the letters express or represent sounds, may be taught to read and write understandingly. […] Hence it would appear that the powers of the letters are not, of necessity, identified with their sounds” (both quotations found in Venezky 1970: 23–24). The second citation is reminiscent of a much earlier paper written by Wallis (on teaching the deaf to speak), where the author claims that “there is nothing, in the nature of the Thing it self, why Letters and Characters might not as properly be applyed to represent Immediately, as by Intervention of Sounds, what our Conceptions are” (Wallis: 1670: 1091). The linguists’ varying attitudes to written language clearly correspond to different expectations of its aims and functions. The early structuralists’ disparaging view of writing has influenced the definitions of orthography, with which writing has commonly been confused (Vachek 1976 [1945–49]: 128). It motivated the treatment of orthography as an imperfect device for obtaining information about speech, and popularized and somewhat fossilized the perception of orthography as a non-linguistic subject. As a result, the English orthographic system has rarely been devoted separate sections, let alone chapters, in historical grammars of English and handbooks of linguistics. Obviously, as it has been impossible to ignore orthography completely, elements of it have been mentioned in discussions of English phonology (see Kniezsa 1991 for an overview of histories of English spelling). Summing up, the main differences between linguists expressing their views about the written language and its connection to speech concern the level of autonomy and the linguistic status of writing systems. Because the purely relational perspective seems biased and limited, in what follows, I have taken the more balanced, autonomistic view towards writing, and consequently towards orthography, but I will also refer to the existing competing views wherever necessary (see, especially, Section 4 on the units of writing systems).
2 Definitions of orthography and related terms According to the OED (Simpson [ed.] 2000–), the word orthography came into English from Old French ortografie, but it originates in Greek. Since its appearance in English in the mid-15th century, it has retained the prescriptive sense of a spelling norm, where spelling is ‘the manner of expressing words with letters’. In the 16th century, it also acquired the descriptive meaning of ‘that part of grammar which treats of the nature and values of letters and of their combination to express sounds and words’. The term orthography has also been defined in various other ways, and its meaning at least partly overlaps with that of other expressions. For example, although in everyday usage orthography is often understood as synonymous only with spelling, in a more technical sense, the orthography of a specific language comprises spelling as well as the capitalization, punctuation, and word division permitted in that language. Spelling is most closely connected with the levels of phonology, morphology, and lexicon, whereas capitalization, punctuation, and word division show some correspondences also with syntax, semantics, and stylistics. Thus, in accordance with this broader sense, one can assume the following definition: “[t]he standardized writing system of a language is known as its orthography” (Crystal 1995: 257).
15 Linguistic Levels: Orthography Another term associated with orthography is graphology, introduced by McIntosh (1961: 107) to denote the study of orthographic systems. The advantage of that term is the parallelism to other linguistic levels, such as phonology and morphology. That term was defined further by Halliday et al. (1964: 50) as “orthography, punctuation, and anything else that is concerned with showing how a language uses its graphic resources to carry its grammatical and lexical patterns”. In this sense, graphology and orthography can also be considered synonymous to the writing system. Graphemics, first recorded in 1951, by analogy to phonemics (Pulgram 1951: 19, see also Stockwell and Barritt 1951 on the relational view of graphemics) is another synonym of orthography. It is defined in the OED as ‘the study of systems of written symbols (letters etc.) in their relation to spoken languages’. However, some linguists have suggested that “the term graphemics should be confined to the study of systems of writing only” (Bazell 1981 [1956]: 68), as well as postulated the introduction of the term graphophonemics for “[t]he discipline concerned with the study of the relationship between graphemics and phonemics” (Ruszkiewicz 1976: 49). The term graphotactics, in turn, usually refers to the syntax of graphemes (understood as units of an orthographic system) (see Haas 1970: 59, Carney 1994: 66–69), defined also as “the laws governing [the] combination of graphemes” (Vachek 1973: 9). All the expressions mentioned above are still found today in descriptions of writing systems, and the choice of particular terminology usually depends on the specific focus and theoretical approach of the writer.
3 Classification of writing systems and principles governing English orthography It is currently assumed that writing systems (or “scripts”) can be classified according to what linguistic level is represented by the written symbols (or graphemes). Thus, the writing systems of natural languages can be divided into “morphographic” and “phonographic” (Rogers 2005: 272). In morphographic systems, the symbols are related primarily to morphemes. Some of those morphemes may constitute words, and therefore such systems have been often referred to as “logographic” (from Greek logos ‘word’, OED; see Sampson 1985; Venezky 1999). Chinese and Sumerian are typical examples of writing systems showing this type of relationship. In phonographic systems, the symbols relate to phonological units. Depending on the type of unit represented, they are divided into “syllabic”, “moraic”, and “phonemic” (or “alphabetic”) (Rogers 2005: 272). In moraic writing, graphemes correspond to morae, where a “mora” is “either a syllableinitial CV sequence or a codal (final) consonant” (Rogers 2005: 250). Moraic systems, such as the Japanese katakana, and Cherokee, have traditionally (though rather imprecisely) been called “syllabic” systems (Sampson 1985; Venezky 1999). In phonemic systems, graphemes represent mainly phonemes, or segments, including consonants and vowels. This type of representation has also been referred to as “alphabetic” (Venezky 1999: 4). The phonemic (or alphabetic) principle implies “biuniqueness”, which “requires not only that a given phoneme is represented by a constant symbol but also that the symbol involved does not represent other phonemes” (Carney 1994: 15; see also Lass and Laing 2007: Section 2.2.1). However, total observance of such one-to-one correspondence between sound and symbol is not recorded among the writing systems of natural languages. Scripts classified as the closest to this principle are classical Greek and Finnish (Rogers 2005: 274).
227
228
II Linguistic Levels The problem with the classification of writing systems is that they are never purely morphographic or phonographic. In fact, the longer a given script is used by a speech community or a nation as an everyday means of communication, the more it evolves, and the more mixed characteristics it acquires (Sampson 1985: 42; Rogers 2005: 272). Examples of taxonomical mixtures “regularly using both morphographic and phonographic symbols” are Egyptian, Maya, and the mixed use of morphographic kanji and two types of phonographic kana in Japanese (Rogers 2005: 272). Even such systems as Sumerian and Chinese, where the morphographic principle is clearly the dominant one, show a limited level of phonography (Sampson 1985: 54; Rogers 2005: 274–275). Writing systems differ not only by their amount of morphography, but also by their “orthographic depth”. The more different morphemes are distinguished in spelling, the higher the level of morphography of a given system (Rogers 2005: 275). For example, English is classified as a phonemic system, but it shows the distinctions among different homophonous morphemes, e.g., you – yew – U – ewe, right – rite – wright – write (Bradley 1904: 214; Rogers 2005: 273; see also Craigie 1928: 1, Bolinger 1946: 335; Vachek 1973: 43–44 for discussions on the distinction between homophones). That property of English spelling has also been referred to as the “lexical principle” (Chomsky and Halle 1968: 49). On the other hand, “[o]rthographic depth is greater if different allomorphs of the same morpheme are written the same” (Rogers 2005: 275). Accordingly, a final in English can be related directly to {past tense}, because it does not show systematic differentiation between its allomorphs /ɪd/ (pointed), /d/ (played) and /t/ (washed). Constant spelling for a morpheme, irrespective of its spoken variants, can also be found in heterophonous, but etymologically and semantically related words, such as child-children, south-southern, and sign-signature (Francis 1958: 562; Venezky 1970: 42–43, 108; Vachek 1973: 25). Such a preservation of morpheme constancy may be considered consistent with the “morphophonemic principle” (Hall 1981 [1960]: 74). Morphemic and morphophonemic spellings breach the “alphabetic principle”, because they lead to a multiplication of sound-symbol (or grapho-phonemic) correspondences, and by the same measure contribute to the opacity, or orthographic depth, of a phonographic system. On the basis of the pairs of etymologically related words provided above, may correspond to /aɪ/ and /ɪ/, to /aʊ/ and /ʌ/, and to /ɡ/ or zero, and these sets are by no means exhaustive lists of all the permissible correspondences. Thus, English orthography can be called “deep”, in contrast to, e.g., the Spanish orthographic system, which shows less divergence from the alphabetic principle, and consequently can be referred to as “shallow”. In fact, “the deep/ shallow contrast is a gradient rather than all-or-none distinction” (Sampson 1985: 45), and morphemic and morphophonemic spelling is not an exceptionless rule in English. Rather, English orthography “represents a level intermediate in depth between the phonemic and the morphophonemic level” (Sampson 1985: 44; see also Hockett 1958: 542). For example, the vowel in the plural suffix (as in pitches) is represented by , but the difference in voicing between /s/ and /z/ (as in cats vs. dogs) is not represented at all. A maximally deep orthography would then have *pitchs instead of pitches. However, the sequence is not permitted in English, and would violate the “graphotactic principle”, determining the permissible letter sequences (Carney 1994: 67). The English writing system also contains a few marginal, non-phonemic, and nonmorphemic, but rather iconic (or pictographic) elements, including, e.g., the symbols ,
15 Linguistic Levels: Orthography and , which represent the words and, at, and the phrase per cent, respectively (Sampson 1985: 34; Lass and Laing 2007: Section 2.2.1). It is noteworthy that the identification of different principles governing the English writing system, and orthographic systems in general, is not a recent discovery. Already in the late 19th and early 20th centuries Jan Baudouin de Courtenay elaborated a set of rules governing orthographic systems. They included three principles: the phonetic, etymological, and historical. Thus, he related orthography to what he believed were its three determining factors, pronunciation, origin (referring to morphology), and tradition (Ruszkiewicz 1981: 24–25; Sgall 1987: 2–3). The idea of multi-level principles and correspondences according to which orthography operates has been further elaborated upon and modified by numerous linguists. For example, Firth (1935: 61) wrote about the notion of “polysystem”, or system of systems, referring to the co-existence of and interaction between phonological, grammatical, and lexical systems of orthographic representation. Overviews of the investigations on the interaction of principles in English orthography and in other orthographic systems can be found, e.g., in Ruszkiewicz (1976), Sgall (1987), and Liuzza (1996). They are summarized by Sgall thus: “In the literature one often speaks about an orthography being based on several principles, the main among which is the phonemic one, while the others underlie the deviations from this basic principle and can be classed more or less exactly in accordance with the levels of the language system” (Sgall 1987: 12).
4 Units of writing systems: terminological evolution In the course of time and investigation into the nature of relationships between written characters and other levels of language, not all linguists have been satisfied with the simple, traditional terms letter and sound, and consequently, there has gradually developed a whole set of terms and definitions in order to make descriptions of the writing system more precise. The terminology used by particular linguists has been developed in close connection with their specific approaches to orthography and their perception of its structure and functions. The following sections discuss some of the most important terms and the ideas behind them.
4.1 The doctrine of littera The most characteristic feature of the early attempts at defining the unit of the writing system is the apparent lack of a clear distinction between letters and sounds. The term letter was used to refer both to written alphabetic marks and to sounds. One can consider, for example, the statement “The Elements of Language are Letters, viz. Simple discriminations of Breath or Voice” (Holder 1669; quoted by Abercrombie 1981 [1949]: 10). What at first sight may seem to be the result of utter confusion, can, in fact, show the impact of the classical notion of littera, introduced by the Stoic grammarians, and described by Aelius Donatus in his Latin grammar in the 4th century CE. Littera referred to the smallest element of language, combining three attributes, nomen (name used for identification), figura (shape or visual configuration), and potestas (power to signify sound). (Abercrombie 1981 [1949]: 14; Henderson 1985: 142). Thus, the visual mark and the sound could be viewed as different aspects of the same entity.
229
230
II Linguistic Levels In spite of its significant and long-lasting influence on numerous generations of grammarians, the concept of littera, in the course of time, proved insufficient for the description of both spoken and written systems of language, at least partly due to its potential for misinterpretation. Quotations such as: “Letters are Signes of Sounds, not the Sounds themselves” (Brightland 1711; quoted by Abercrombie 1981 [1949]: 11) demonstrate that the complexity of the concept of littera was not transparent and could lead to ambiguity. However, after years of abandonment in the 20th century, the notion has recently been revived by linguists and it has inspired the development of such concepts as “litteral substitution sets” (LSS), which have proved useful in the description of early Middle English writing systems (see, e.g., Benskin 1991: 226; Laing 1999; Laing and Lass 2003, 2009; Lass and Laing 2007: Sections 3.3.1 and 2.3.2). The word letter is still in use both in everyday language with reference to a character of an alphabet, or in more technical descriptions, usually as part of a definition of the distinctive unit of the orthographic system.
4.2 Grapheme Since the introduction of the term grapheme by Baudouin de Courtenay in 1901 (Ruszkiewicz 1976: 24–37, 1981 [1978]: 20–34), it has been defined in various ways: – “the class of graphs which denote the same phoneme” (Hammarstro¨m 1981 [1964]: 97); – “the class of letters and other visual symbols that represent a phoneme or cluster of phonemes” (OED, Simpson [ed.] 2000–); – “any minimal letter string used in correspondences” (Carney 1994: xxvii); – “[t]he unit of writing” (Stetson 1981 [1937]: 35); – “the minimal functional distinctive unit of any writing system” (Henderson 1984: 15); and – “a purely distinctive visual unit, part of an autonomous semiotic system” (Liuzza 1996: 28). As can be seen from these quotations, the available definitions can be divided into two groups, corresponding to two main senses, and reflecting “conflicting linguistic views of the status of writing” (Henderson 1985: 142): 1. a letter or cluster of letters referring to or corresponding with a single phoneme; 2. the minimal distinctive unit of a writing system. The former sense, evident in the first three quotations above, assumes that orthography is mainly, if not exclusively, a means of notation for representing speech, and is usually adopted by the proponents of the relational approach to orthography. The latter sense, in turn, is more typical of the adherents to the autonomistic approach to writing (see Section 1 above). It emphasizes the contrastiveness of the grapheme within the orthographic system as one of its necessary features. Moreover, definitions associated with this sense also occasionally point to the analogy which can be drawn between the status of graphemes and that of phonemes within their respective systems: “the graphemes of a given language – like its phonemes – remain differentiated from one another, i.e. […]
15 Linguistic Levels: Orthography they do not get mixed up” (Vachek 1976 [1945–49]: 128–129). Pulgram (1951: 15) even proposed a comprehensive list of correspondences between the graphemes and phonemes of a language. He was also one of the first linguists to use the term allograph, by analogy to allophone. According to him, “all graphs identifiable as members of one grapheme are its allographs”. This definition was later developed by McLaughlin: “an allograph or allographic set which contrasts significantly with all other allographs or allographic sets or with zero will be called a GRAPHEME”, where an allograph is “a group of similar characters, modifications, or features [i.e. graphs] classed together […] in graphemic analysis” (McLaughlin 1963: 29). Within the graphemic system, graphs can be defined as “each hic et nunc realization of a grapheme” (Pulgram 1951: 15). It was cogently argued by Henderson (1985) that the autonomistic definition of the grapheme, in comparison to the relational one, “lends itself to the most coherent and principled use. It also allows the term to be applied across the range of writing systems, including irregular alphabetic systems, syllabaries and logographic systems” (Henderson 1985: 146).
5 The inventory and distribution of English graphemes from the synchronic and diachronic perspective The inventory of graphemes used in modern English comprises “alphabetic” and “nonalphabetic” graphs. The former refer to the twenty-six letters of the alphabet . Each letter has also a capitalized counterpart, e.g., , so altogether we can distinguish 52 alphabetic characters. Opinions about the graphemic significance of capital letters are divided. According to Haas (1970: 22–23), since the lower-case and upper-case alphabetic graphs occur in complementary distribution, they can be described as allographs of a particular grapheme, e.g., or . Taking this point of view, we could distinguish 26 (not 52) graphemes in English. However, Henderson points to the fact that in sentences such as “The archer was called Archer” archer and Archer are different lexical identities and “so the contrast expressed by capitalization has graphemic status” (Henderson 1985: 144). The non-alphabetic graphic symbols appear in combination with the alphabetic ones, and are subdivided into four types (McLaughlin 1963: 30): 1. “punctuation marks”, usually helping to indicate grammatical structure, including, e.g., < , . : ; ’ ? ! - – … ‘ ’ “ ” ( ) < > [ ] { } >; 2. “graphic components” referring to diacritics, conjoined to other graphs and used to distinguish different values of the same letter, including accent marks, dots, cedillas, subscript hooks, macrons, e.g. (these appear in English only occasionally, in borrowings); 3. “tachygraphs” (found mainly in medieval manuscripts), used only with alphabetic symbols, and standing for one or more of them, e.g., a bar through the descender of the letter
, functioning as an abbreviation for or in, e.g., person; and 4. “word signs”, e.g. (an ampersand), standing for the word and, or the symbol , which stands for at. Moreover, Francis (1958: 436) suggests including a space in the writing system, as a sort of zero grapheme (see also Venezky 1970: 47 and Carney 1994: 5, cf. Sgall 1987: 7 who uses the expression “blank”).
231
232
II Linguistic Levels Viewed from a diachronic perspective, the inventory of English graphemes should also include several additional symbols, particularly thorn , eth , wynn , ash , and yogh . Thorn and wynn were “borrowed” from the runic alphabet (futhorc) in medieval times. Ash was a ligature derived from the Roman alphabet, eth was an Irish modification of to represent, along with thorn, the dental fricatives, and yogh originated as the insular open g (ᵹ), borrowed from the Irish. All these characters had been replaced by other symbols by modern times. On the other hand, the differentiation between and , as well as between and as separate letters of the alphabet is a comparatively new invention in the English inventory of graphemes. (OED, Simpson [ed.] 2000–; Scragg 1974: 2, 8, 10 passim). Over the centuries, changes affected not only the composition of the inventory of symbols used for writing in English. Graphotactics and grapho-phonemic correspondences have also evolved over time, making English orthography more and more inconsistent with the alphabetical principle governing phonemic writing systems. For example, the Old English corresponded to /ʧ/, and /k/, but never to /ʃ/, and corresponded to /ʃ/, but never to /sk/. Likewise, punctuation, capitalization, and word division patterns have been modified over the centuries. The reasons for the changes have been multiple and diverse, but a few important events and influences are particularly noteworthy in the development of English graphemics. For example, the arrival of Christianity at the end of the 6th century CE brought the adoption of the Latin alphabet (with some modifications), and the Norman Conquest resulted in the appearance of numerous French-derived spelling conventions, which partly replaced the previous practices. For example, the Norman scribes introduced the doubling of letters to indicate long vowels in words such as goose and meet, and also replaced , , , and with , , , and , respectively. The introduction of printing contributed to the fixing and standardization of English spelling, or at least to the dissemination of standardized practices; and extensive borrowing from various languages explains the diversity of spelling patterns. Readers interested in more detail on the structure and the rules governing English orthography, can consult a number of books outlining the synchronic and diachronic aspects of English orthography. The most recent synchronic summaries include Rollings (2004), Carney (1994), Venezky (1970, 1999), and Wełna (1982). However, the only relatively reliable comprehensive study devoted to the history of English spelling published so far is Scragg (1974).
6 Orthography as the source of phonological evidence In view of the depth of English orthography, due to the variety and complexity of principles governing it (discussed in the previous sections), a linguist studying the phonology of an older text cannot take orthographic evidence at face value as the source of information about pronunciation, while one studying the morphology or syntax will find relevant evidence much easier to obtain from a written text. Penzl (1957: 197–200) provides a classification of the types of evidence available for students of previous pronunciation, dividing it into orthographic, orthoepic (referring to the comments of phoneticians, grammarians, and spelling reformers), metrical, comparative, and contact evidence. The evaluation of various sources of information about the
15 Linguistic Levels: Orthography phonemic change can be found, apart from Penzl (1957), also in Ko¨keritz (1953), Dobson (1957–68), Wrenn (1967), Wolfe (1972), Liuzza (1996), Stenroos (2002, 2004, 2006), and in earlier works, e.g., Zachrisson (1913) and Wyld (1936). Orthography is thus one of several sources of evidence for historical phonologists, and should at best be analyzed in conjunction with the information gathered from the examination of other types of evidence. Orthographic evidence comprises occasional spellings (sometimes referred to as naive spellings), e.g. douter for daughter, or ruff for rough, and back spellings (also called reverse spellings or hypercorrections), e.g. quight for quite (by analogy to light, night). Unfortunately, the interpretation of such orthographic evidence is fraught with problems: 1. not every change in spelling means a change in pronunciation; e.g. the etymologizing insertion of in doubt did not result in assigning any sound value to that letter; 2. not all changes in pronunciation are reflected in spelling; e.g. the diphthongization of Middle English [iː] and [uː], as in mice and house, is not; 3. occasional spellings may be due to (typo)graphical errors, especially if one deals with a single instance of a given form; and 4. the value of the same spellings may vary according to the period and regional dialect. Nevertheless, a linguist can minimize the risks connected with such types of evidence by gathering a large number of examples before drawing conclusions, by comparing orthographic evidence with other types of evidence, and, most importantly, by examining thoroughly the orthographic rules of the relevant writing system with due attention paid to the period and regional variety, before starting the analysis (see, e.g., Penzl 1957: 197–203; Wolfe 1972: 110–130; Stenroos 2002: 445–468; Smith 2006: 136). Liuzza (1996: 33) emphasizes particularly the last measure mentioned above: “Only by reconstructing the orthographic rules practiced by a writer can one determine anything about the spoken language behind the text; the evidence, otherwise, is mute”.
7 Sociolinguistic aspects of orthography As has been indicated in the previous sections, orthography can be treated as one of the subjects of linguistic research. Nevertheless, in the estimation of an average language user, orthography functions nowadays mainly as a spelling norm. Such a norm is binding within a language community and non-compliance with it can have social consequences. It applies particularly to public domains, such as education and official written media. Someone using non-standard forms and making spelling mistakes in these contexts is likely to be socially punished by being judged uneducated or even unintelligent. The notion of spelling error is a relatively recent development. In the Middle Ages, there was no orthographic standard in English in the modern sense. Instead, dialectal variation was abundant. In the 15th century the situation started to evolve towards standardization. This process took more than two centuries to complete, including the stages of selection, acceptance, functional elaboration, and codification (Haugen 1972: 110). The process of codifying those orthographic forms which were viewed as proper was prompted and advocated by the spelling reformers, grammarians, and lexicographers. Their proposals were essential for the regularization of spelling in English (Brengelman 1980). Spelling books, grammars, and dictionaries, which were printed and
233
234
II Linguistic Levels disseminated by a growing number of publishing houses, contributed to the elimination of most orthographic variants and to the considerable increase in orthographic consistency. This increase was so substantial that it is usually recognized that by the end of the 17th century English orthography had become standardized (Scragg 1974: 80, Go¨rlach 2001: 78). However, it has recently been emphasized that the completion of standardization can be applied mainly to printed documents, whereas “[d]eviant spelling continued in letters and diaries, even among the educated” (Go¨rlach 2001: 78; see also Salmon 1999: 44; Tieken-Boon van Ostade 2009: 46–50). Furthermore, also in the 19th-century English, “while public printed texts manifest greater stability, even these are not devoid of change” (Mugglestone 2006: 278). In present times, with well-established orthographic standards, orthographic variation still exists. In English, there is the codified and politically sanctioned diatopic division into two spelling systems, British and American, including the respective use of such spelling variants as those in honour/honor, centre/center, monologue/monolog, and aeroplane/ airplane. The spelling can differ also depending on the particular reference of the word in a given context. For example, in British English, the form program is reserved for computingrelated contexts, and programme is used in the other meanings, whereas American English uses program in all senses of that lexical item (OED, Simpson [ed.] 2000–). In today’s languages, the use of non-standard orthographic variants can also be deliberate (Beaugrande 2006: 43), or even constitute “a powerful expressive resource” capturing some of the “immediacy, the ‘authenticity’ and ‘flavor’ of the spoken word in all its diversity” (Jaffe 2000: 498). One can experience the effectiveness of non-standard orthography in daily contexts, for example in trade-names, television commercials and in computer-mediated communication (Crystal 2001) (cf. Heyd, Chapter 70). It can also be found in printed texts, for instance in poetry (e.g. Tom Leonard), or popular fiction (e.g. Irvine Welsh), where it is used for stylistic as well as ideological reasons.
8 Summary Orthography differs in many ways from the other levels of linguistic description. It does not enjoy a stable and generally accepted terminology, partly because in numerous natural languages, such as in English, it is governed by a set of heterogeneous principles, and partly due to the fact that linguists differ significantly as regards the aims and functions which they ascribe to orthographic systems. The existence of diverse definitions of the term orthography itself reflects the complexity typical of writing systems. This term is employed to denote, on one hand, a system of graphemes as contrastive units, and, on the other, a set of rules governing the correspondences between the graphemes, phonemes, and morphemes. It also functions as a social code which needs to be observed under the pain of stigmatization. In the outline offered above, I have attempted to present some of the important aspects of the controversy concerning the relationship between speech and writing, discuss the different levels of linguistic representation in orthography, and describe the main principles according to which the English writing system operates. The characteristics of this system are adequately expressed by Venezky (1999: 4): “English orthography is not a failed phonetic transcription system, invented out of madness or perversity. Instead, it is a more complex system that preserves bits of history (i.e. etymology), facilitates understanding, and also translates into sound”.
15 Linguistic Levels: Orthography
9 References Abercrombie, David. 1981 [1949] What is in a “letter”? In: Ruszkiewicz (ed.), 9–19 [Reprinted from Litera 3: 43–46]. Bazell, Charles E. 1981 [1956] The grapheme. In: Ruszkiewicz (ed.), 66–70 [Reprinted from Lingua 2: 54–63]. de Beaugrande, Robert. 2006. Speech versus writing in the discourse of linguistics. Miscela´nea: A Journal of English and American Studies 33: 31–45. Benskin, Michael. 1991. In reply to Dr Burton. Leeds Studies in English: New Series 22: 209–262. Bloomfield, Leonard. 1933. Language. New York: Holt, Rinehart, and Winston. Bolinger, Dwight. 1946. Visual morphemes. Language 22: 333–340. Bradley, Henry. 1904. The Making of English. London: Macmillan. Bradley, Henry. 1919 [1913] On the Relations between Spoken and Written Language, with Special Reference to English. Oxford: Clarendon Press. [Reprinted] Brengelman, Fred H. 1980. Orthoepists, printers, and the rationalization of English spelling. Journal of English and Germanic Philology 79: 332–354. Carney, Edward. 1994. A Survey of English Spelling. London/New York: Routledge. Chomsky, Noam and Morris Halle. 1968. The Sound Pattern of English. New York: Harper and Row. Cook, Vivian. 2004. The English Writing System. London: Hodder Arnold. Craigie, William A. 1928. English Spelling: Its Rules and Reasons. London: George G. Harrap. Crystal, David. 1995. The Cambridge Encyclopedia of the English Language. Cambridge: Cambridge University Press. Crystal, David. 2001. Language and the Internet. Cambridge: Cambridge University Press. Dobson, E. J. 1957–68 . English Pronunciation: 1500–1700. Vols. I and II. Oxford: Clarendon Press. Firth, John R. 1935. The technique of semantics.Transactions of the Philological Society, 36–72. Francis, W. Nelson. 1958. The Structure of American English. New York: The Ronald Press Company. Go¨rlach, Manfred. 2001. Eighteenth-century English. Heidelberg: C. Winter. Haas, William. 1970. Phono-graphic Translation. Manchester: Manchester University Press. Hall, Robert A. 1981 [1960] A theory of graphemics. In: Ruszkiewicz (ed.), 71–80 [Reprinted from Acta Linguistica 8: 13–20]. Halliday, Michael A. K., Angus McIntosh, and Peter Strevens. 1964. The Linguistic Sciences and Language Teaching. Bloomington: Indiana University Press. Hammarstro¨m, Go¨ran. 1981 [1964] Type and typeme, graph and grapheme. In: Ruszkiewicz (ed.), 89–99 [Reprinted from Studia Neophilologica 36: 332–340]. Haugen, Einar. 1972. Dialect, language and nation. In: John Pride and Janet Holmes (eds.), Sociolinguistics, 11–24. Harmondsworth: Penguin Modern Linguistics Readings. Henderson, Leslie. 1984. Writing systems and reading processes. In: Leslie Henderson (ed.), Orthographies and Reading: Perspectives from Cognitive Psychology, Neuropsychology, and Linguistics, 11–24. London: Lawrence Erlbaum. Henderson, Leslie. 1985. On the use of the term “grapheme”. Language and Cognitive Processes 1(2): 135–148. Hockett, Charles F. 1958. A Course in Modern Linguistics. New York: The Macmillan Company. Jaffe, Alexandra. 2000. Introduction: Non-standard orthography and non-standard speech. Journal of Sociolinguistics 4(4): 497–513. Kniezsa, Veronika. 1991. “The due order and reason”: On the histories of English spelling. Folia Linguistica Historica 12(1–2): 209–218. Ko¨keritz, Helge. 1953. Shakespeare’s Pronunciation. New Haven: Yale University Press. Laing, Margaret. 1999. Confusion wrs confounded: Litteral substitution sets in early Middle English writing systems. Neuphilologische Mitteilungen 100: 251–270. Laing, Margaret and Roger Lass. 2003. Tales of the 1001 nists: The phonological implications of litteral substitution sets in some thirteenth-century South-West Midland texts. English Language and Linguistics 7: 257–278.
235
236
II Linguistic Levels Lass, Roger and Margaret Laing. 2007. Introduction. Part I: Background. Chapter 2: Interpreting Middle English. In: Margaret Laing and Roger Lass, LAEME: A Linguistic Atlas of Early Middle English, 1150–1325. University of Edinburgh. http://www.lel.ed.ac.uk/ihd/laeme1/laeme1. html (date of access: 18 November 2009) Laing, Margaret and Roger Lass. 2009. Shape-shifting, sound-change and the genesis of prodigal writing systems. English Language and Linguistics 13(1): 1–31. Liuzza, Roy M. 1996. Orthography and historical linguistics. Journal of English Linguistics 24(1): 25–44. McIntosh, Angus. 1961. “Graphology” and meaning. Archivum Linguisticum 13: 107–120. McLaughlin, John C. 1963. A Graphemic-phonemic Study of a Middle English Manuscript. The Hague: Mouton. Mugglestone, Lynda. 2006. English in the nineteenth century. In: Mugglestone (ed.), 274–304. Mugglestone, Lynda (ed.). 2006. The Oxford History of English. Oxford: Oxford University Press. Penzl, Herbert. 1957. The evidence for phonemic changes. In: Ernst Pulgram (ed.), Studies Presented to Joshua Whatmough on his Sixtieth Birthday, 193–208. The Hague: Mouton. Pulgram, Ernst. 1951. Phoneme and grapheme: A parallel. Word 7: 15–20. Rogers, Henry. 2005. Writing Systems: A Linguistic Approach. Malden: Blackwell. Rollings, Andrew G. 2004. The Spelling Patterns of English. Mu¨nchen: Lincom. Ruszkiewicz, Piotr. 1976. Modern Approaches to Graphophonemic Investigations in English. Katowice: Uniwersytet S´la˛ski. Ruszkiewicz, Piotr. 1981 [1978] Jan Baudouin de Courtenay’s theory of the grapheme. In: Ruszkiewicz (ed.), 20–34 [Reprinted from Acta Philologica 7: 117–135]. Ruszkiewicz, Piotr (ed.). 1981. Graphophonemics: A Book of Readings Katowice: Uniwersytet S´la˛ski. Salmon, Vivian. 1999. Orthography and punctuation. In: Roger Lass (ed.), The Cambridge History of the English Language. Vol. III. 1476–1776, 13–55. Cambridge: Cambridge University Press. Sampson, Geoffrey. 1985. Writing Systems: A Linguistic Introduction. Stanford: Stanford University Press. Sapir, Edward. 1921. Language: An Introduction to the Study of Speech. New York: Harcourt, Brace & World. de Saussure, Ferdinand. 1993. Troisie`me Cours de Linguistique Generale (1910–1911): d’apre`s les cahiers d’Emile Constantin./Third Course of Lectures on General Linguistics (1910–1911): From the Notebooks of Emile Constantin. Eisuke Komatsu (ed.) and Roy Harris (trans.). Oxford: Pergamon Press. Scragg, Donald G. 1974. A History of English Spelling. New York: Manchester University Press and Barnes & Noble. Sgall, Petr. 1987. Towards a theory of phonemic orthography. In: Philip A. Luelsdorff (ed.), Orthography and Phonology, 1–30. Amsterdam/Philadelphia: John Benjamins. Simpson, John (ed.). 2000–. The Oxford English Dictionary. 3rd edn. online. Oxford University Press. www.oed.com/ Smith, Jeremy J. 2006. From Middle to Early Modern English. In: Mugglestone (ed.), 120–146. Stenroos, Merja. 2002. Free variation and other myths: Interpreting historical English spelling. Studia Anglica Posnaniensia 38: 445–468. Stenroos, Merja. 2004. Regional dialects and spelling conventions in Late Middle English: Searches for (th) in the LALME data. In: Marina Dossena and Roger Lass (eds.), Methods and Data in English Historical Dialectology, 257–285. Bern: Peter Lang. Stenroos, Merja. 2006. A Middle English mess of fricative spellings: Reflections on thorn, yogh and their rivals. In: Marcin Krygier and Liliana Sikorska (eds.), To Make his Englissh Sweete upon his Tonge, 9–35. Frankfurt a. Main: Peter Lang. Stetson, Raymond H. 1981 [1937] The phoneme and the grapheme. In: Ruszkiewicz (ed.), 35– 44 [Reprinted from Me´langes de linguistique et de philologie offerts a Jacq. van Ginneken, 353–356].
16 Linguistic Levels: Styles, registers, genres, text types
237
Stockwell, Robert P. and C. Westbrook Barritt. 1951. Some Old English Graphemic-Phonemic Correspondences – ae, ea and a. (Studies in Linguistics, Occasional Papers, 4.) Norman, OK: Battenburg Press. Tieken-Boon van Ostade, Ingrid. 2009. An Introduction to Late Modern English. Edinburgh: Edinburgh University Press. Vachek, Josef. 1973. Written Language: General Problems and Problems of English. The Hague: Mouton. Vachek, Josef. 1976 [1945–49] Some remarks on writing and phonetic transcription. In: Vachek (ed.), 127–133 [Reprinted from Acta Linguistica 5: 86–93]. Vachek, Josef (ed.). 1976. Selected Writings in English and General Linguistics. The Hague: Mouton. Vachek, Josef. 1982. English orthography: A functional approach. In: William Haas (ed.), Standard Languages: Spoken and Written, 37–56. Manchester: Manchester University Press. Venezky, Richard L. 1970. The Structure of English Orthography. The Hague: Mouton. Venezky, Richard L. 1999. The American Way of Spelling. New York: The Guildford Press. Wallis, John. 1670. A Letter of Doctor John Wallis to Robert Boyle Esq. concerning the said Doctors Essay of Teaching a person Dumb and Deaf to speak, and to understand Language, together with the success thereof, made apparent to his Majesty, the Royal Society, and the University of Oxford. Philosophical Transactions 61: 1087–1099. Wełna, Jerzy. 1982. English Spelling and Pronunciation. Warszawa: Pan´stwowe Wydawnictwo Naukowe. Wolfe, Patricia. 1972. Linguistic Change and the Great Vowel Shift in English. Berkeley: University of California Press. Wrenn, Charles L. 1967. The value of spelling as evidence. In: Charles L. Wrenn (ed.), Word and Symbol: Studies in English Language, 129–149. London: Longmans. Wyld, Henry. 1936. A History of Modern Colloquial English. Oxford: Basil Blackwell. Zachrisson, Robert. 1913. Pronunciation of English Vowels 1400–1700. Go¨teborg: W. Zachrissons boktryckeri.
Hanna Rutkowska, Poznan´ (Poland)
16 Linguistic Levels: Styles, registers, genres, text types 1 2 3 4 5 6
Introduction A selective history of registers Stylistic developments in English Approaches to historical texts Summary References
Abstract The chapter surveys research on registers, styles, text types and genres in the history of English. The presence of registers is connected with socio-cultural conditions, such as the structure of society, multilingualism, the practice of translation, academic traditions, and technological progress. The legal and the scientific registers are described in more detail. The stylistic development of English is treated here as linked to standardization Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 237–253
238
II Linguistic Levels (elaboration in syntax and lexicon) and the orality-literacy continuum (increasing development of more literate characteristics). The curial style of the 14th/15th century and the plain style of the 17th century are highlighted. Three approaches to genre and to text types are presented. Inventories of genre labels highlight the presence, nature and development of genres throughout English history. Changes in texts, and genres across periods can also be studied through looking at their conventional structure (e.g. letters) or their linguistic features (e.g. discourse deixis). Lastly, text types/genres can also play a role in language change.
1 Introduction Looking through books on the history of English, one will as a rule not find separate sections on texts and registers (with the exception of Go¨rlach 1999a, 2001a). Looking at language in use, in contrast, it is obvious that it “exists in texts” (Diller 2001: 3). Individual linguistic features are realized in texts and for the sake of creating coherent and effective texts, so that language history can not only be investigated by studying texts but is actually only constituted by texts and changing textual needs. Any linguistic history that is textless, therefore, ignores a very important perspective on the development of English. The inter-relationship between the development of linguistic features and their textual uses has only relatively recently received more attention on a firmer theoretical and methodological basis than before. The collection edited by Diller and Go¨rlach (2001) and the special issue of the European Journal of English Studies (ed. Moessner 2001) dealing with genre and text types in a historical perspective bear witness to this. Unfortunately, the definitions of the terms in the title are quite varied in the literature (cf. Diller 2001 for a discussion). This article is based on the following understanding of the terminology. While registers and styles are inventories of linguistic devices, genres and text types are classes of texts. Genre is primarily based on text-external considerations (cf. Biber 1989: 5–6) and refers to aspects such as the functions, conventional shape, and structure of texts. Genres are linked to expectations on the part of text users about the (proto-)typical functions and (surface) features of texts belonging to the genre; thus competent speakers have fairly clear ideas about what a fairy tale, a letter, a prayer, or a weather forecast is like. As folk categories, genres are not necessarily defined by a strict and homogeneous set of criteria. Text types, on the other hand, are defined by text-internal linguistic criteria, which to a certain extent go along but do not completely overlap with genre distinctions (cf. Taavitsainen 2002: 220). Texttypological approaches (e.g. Longacre 1996; Werlich 1983) have presented broad categories such as narrative, descriptive, expository, instructive, or procedural/behavioral, and argumentative, which are characterized by a typical (co-)occurrence of linguistic features, e.g. narrative: past tense verbs, time adverbials; procedural: imperatives. Such internal text-type features can be present in a given text to varying extents, thus making it more or less expository/narrative/etc. in nature, and features from different types can combine, producing, for example, an expository-argumentative hybrid. Register is a more general term, comprising both oral and written productions based in particular on situational, social, and professional contexts and the field or domain of discourse (cf. also Lenker, Chapter 21). Thus the domains of religion, law, science, journalism, etc., constitute the religious, legal etc. registers, all of which exhibit a certain cohesion in terms of possible interaction types, aims, and contents, producing
16 Linguistic Levels: Styles, registers, genres, text types lexico-grammatical similarities on a more general level than text types. Registers usually comprise various genres. Style is the vaguest of all these terms and potentially cuts across all the other distinctions. Diller (1998: 155–156) defines style as used by textual stylistics as the idiosyncractic “characteristic linguistic features of a text”, and that of linguistic stylistics as “different ways of saying the same thing”. From the point of view of the language user, style implies aspects like choice between linguistic items, perceptions of appropriateness going beyond register conventions, ideas of norm vs. deviance, and (potentially prescriptive) aesthetic notions. A given style can go with an individual, a group of people, or a time period. A note on the application of some of these concepts in a commonly used source in historical linguistics, namely the Helsinki Corpus (HC) (Rissanen et al. 1991), may be in order here. The corpus encodes what it calls “text type” () and “prototypical text category” (). The latter refers to (broad) text types as explained above, containing such labels as “expository”, “instruction religious”, and “narration imaginative”, to which it adds register information (religious) or the more general distinction of (non-)fictionality in some cases. Helsinki Corpus text types, in contrast, in some cases are rather genres (e.g. handbook, sermon, preface, comedy) and in others are reminiscent of registers (e.g. science medicine, history). The emphasis of this article will be on non-literary texts (cf. Fulk, Chapter 25, Arnovick, Chapter 35, and Moore, Chapter 50). There will be some overlap with the notion of discourse as treated in Lenker, Chapter 21.
2 A selective history of registers Some registers are attested throughout the history of English, albeit perhaps in varying strength and internal variety. The religious register is one of those with a long history, for example, and also with a certain breadth of representation. Apart from the Bible (cf. Kohnen, Chapter 65), the register is represented by – both original and translated – sermons and homilies, texts related to the liturgy like prayers, the Creed, the psalter or hymns, lives of saints, hortative and instructional writing, and academic theological/exegetical writing. Except for the last type, all are attested from Old English onwards. Other registers, in contrast, have only emerged at some later point in history (e.g. newspaper language, cf. Fries, Chapter 67). Late evolution may have to do with extralinguistic developments, such as the possibilities offered by the printing press or the rise of modern natural science, but also with the different status of English vis-a`-vis other languages in different periods. As to the latter aspect, in domains like religion, the law, and scientific/academic writing, Latin in particular, and French played important roles during the Middle Ages and also beyond, thus competing with, sometimes dominating, English in the respective registers (cf. Go¨rlach 1999b: 462). This will also be visible in the short outline of the registers of law and science which follows.
2.1 Legal English Like the religious register, legal (and administrative) English is a long-standing variety with a remarkable functional stability and a very distinctive form of English. Hiltunen (1990) provides a historical survey of legal English, but without covering all periods equally (Early Modern English is particularly neglected). The oldest extant legal
239
240
II Linguistic Levels texts are the laws of King Ethelbert of Kent (635 CE), which are followed by various other law codes up to the reign of Cnut in the 11th century (cf. Liebermann 1903–16). Old English legal language already shows the complex structures present in the modern variety, even if to a lesser extent, such as conditional, relative and adverbial subordination, and multiple embedding. But it also lacks some modern characteristics, such as the emphasis on unambiguous reference and precision, and thus also its repetitiveness. With the Norman Conquest, however, English ceased to exist as a language of the law for about four centuries, until the 1362 Statute of Pleading re-established English as the oral legal language, and the first Act of Parliament to be written in English was passed in 1483. During the Middle English (ME) period, legal writing had used first “Law Latin”, later French, while pleading had taken place in French, thus adopting not only foreign legal procedures, but also the linguistic patterns of legal Latin and French. Similar linguistic mixtures are also found in administrative or business records, whose macaronic writing style (mixing Latin, French, and English) has been termed a “deliberate, formal register” by Wright (1992: 769). The full establishment of English in all spheres of law was gradually carried through during the Early Modern English period, involving again translation of important texts into English. From about the 16th century there is no shortage of legal texts in English. Through trial transcripts and proceedings we also have an insight into the oral forms of historical legal discourse; studies on Early Modern English courtroom language have been carried out by Archer (2005) on British data and by, e.g., Grund (2007) on the Salem Witchcraft Trials. As a consequence of its history, Hiltunen (1990: 52) has characterized modern legal language as “essentially a kind of ‘creole’, where the formative elements go back to an amalgamation of native resources and extensive borrowing”. One of the noteworthy characteristics of legal English is its lexicon reflecting the influence of the various legal traditions throughout its history, from Anglo-Saxon terminology (which has mostly disappeared, often together with the concepts, e.g. wergeld), via Norse terms, to French and Latin words (Mellinkoff 1963). Modern law language is further characterized by archaic lexical usages, such as aforesaid, theretofore and similar elements, which are fossils from the Early Modern English period. Throughout its English history, the register seems to have followed a trend towards ever more specificity and explicitness, thus increasing its linguistic complexity. Partly, this complexity has been made more accessible by structure-building visual arrangements, which had not been used in the past; this last point shows that the visual aspect of texts should also play a role in textual studies.
2.2 Scientific language The “vernacularisation boom of the fourteenth century” (Taavitsainen 2001: 189) was certainly important for law, and so it was for another register, i.e. scientific writing. In contrast to law, science is a younger discipline and thus also a newer register. Anglo-Saxon science and an Old English scientific register in a proper sense do not exist; what is extant is rather texts of a practical nature, such as astrological texts, herbals, and medical recipes (e.g. Bald’s Leechbook). In general the scientific or academic community of the Middle Ages, which was an international one, used Latin, not only as a written but also as a spoken language, e.g. within the universities. This custom was
16 Linguistic Levels: Styles, registers, genres, text types slow to die, extending into the 17th and 18th centuries in England, when Latin works were still being written, although to an ever-decreasing extent. From an extralinguistic point of view, science – in the modern understanding of an empirical (also experimental) and rationalist undertaking in search of new knowledge, which is prototypically represented by the natural sciences – is a product of the 17th and 18th centuries (Hunter 1981). What we find before that is the more traditional and conservative scholastic tradition as well as various precursors of modern science from the late Middle Ages onwards, both with regard to thinking, method, and language use. This development is being charted by the “Scientific Thought-styles” project centering on the exemplary investigation of medical writing, which revealed a shift from argumentation founded on established authorities to more evidence-based argumentation. The split of the domain into a learned-popular continuum, emerging from late ME (Taavitsainen 2005), can be seen in this connection; while the learned end is characterized by expository and argumentative texts, the popular side tends to be practically oriented and instructive in style. Taavitsainen (2001) presents a survey of the early development of the scientific register. Vernacular writing in medicine, followed by other scientific fields (e.g. astronomy), emerged in the 14th century. Some forms of writing in this field had a continuous vernacular history, e.g. recipes, rules of health, charms, prognostications, and remedy books (Taavitsainen 2005). Generally, however, the developing English register took as its model the Latin academic and scientific register (termed “modelling from above” by Taavitsainen 2001: 188), which was aided by the fact that many texts were either translations (e.g. by Trevisa in the 14th century) or adaptations of foreign sources. Newly introduced genres, such as specialized treatises or surgical texts, often made their way into English through translations. In particular the conventions for the learned genres were apparently taken over from the Greco-Roman tradition, but research in this area is complicated by the fact that the relevant genres in Latin and Greek have not been sufficiently investigated (Taavitsainen 2005). Taavitsainen mentions that the classical format of questions and answers, which was simplified in English, developed from fairly irregular early attempts to fixed and regular structures in the 16th century, and remained in use in handbooks for centuries. The scientific vocabulary is thus also based on extensive borrowing from Romance sources (increasing from the 14th century onwards), with the respective greater use of native vs. foreign lexical resources corresponding to more practical/popular writings vs. more theoretical/academical writing in the medical field, for example. While the Helsinki research group around Taavitsainen and Pahta concentrates on medical writing, Halliday (1988) charted the development of physics writing as a representative of the scientific register. He shows how the major propositional points are over time increasingly presented in nominalized forms, easily allowing both objectivization, categorization, and fore-/backgrounding, while verbs are progressively more restricted to expressing relational or existential aspects of the nominal arguments. In parallel, though starting somewhat later than the just mentioned aspects, the depersonalisation of scientific discourse is in progess. Let me end this section with two general points. First, translations have played a role in various registers. This means that register and textual studies will have to pay attention to how foreign models influenced English developments, and thus, more crosslinguistic studies and also more text-oriented studies on languages such as Medieval Latin are necessary. Second, a link can be made between register development and
241
242
II Linguistic Levels standardization in English (cf. also Section 3), as different registers had a differing impact on the standard and also were receptive to the standard to various extents. The administrative register in the form of Chancery writings was one input into the emerging English standard (Fisher 1996), Lollard texts within the religious register have been connected with another important strand of standardizing varieties (Samuels 1963), with which ME scientific writing also had certain affinities, although this register seems to have resisted the standard somewhat more (Taavitsainen 2001). In EModE times, the Book of Common Prayer and the Authorized Version also had a standardizing impact.
3 Stylistic developments in English Styles are at least partly linked to the orality-literacy continuum, with oral and literate features mixing in particular ways in texts. In this connection, standardization, especially the process of elaboration (Haugen 1972), is important for stylistic developments, as standardization will emphasize the literate end of the continuum. Another important aspect is the varying impact of foreign models, French and in particular Latin, which goes together with the long-standing but ultimately waning influence of (classical) rhetoric. Furthermore, specific stylistic changes are sometimes embedded in certain registers and genres, literary as well as non-literary, which also accounts for the substantial contributions to historical stylistics from literary studies. A last aspect of note is the fact that most studies of style deal with the (early) modern period, at the most extending into late Middle English. Gordon (1966), still the only large-scale historical study of prose style extant, shows this bias as well, the Middle Ages being dealt with in a mere 35 pages.
3.1 Standardization processes and their stylistic effects The syntactic elaboration of English proceeding especially during the 16th to 18th centuries ultimately produced more complex, ordered, and explicit structures, all of them principally useful for complex writing and thus essentially literate (e.g. Rissanen 1999 for an overview of syntactic developments). Such stylistically relevant structures include greater and more sophisticated use of hypotactic and embedded structures (adverbial and relative clauses, non-finite constructions, innovations in the conjunction class) and emerging and/or expanding use of topicalizing constructions (e.g. passive, clefting), as well as more discriminate use of prepositions and prepositional phrases, enabling a tighter information structure. In Chafe’s (1982) terms, these characteristics make for the more integrated and detached characteristics of modern writing. Despite the fact that it is difficult to clearly attribute individual changes to foreign, especially Latin, models, the general attitude towards Latin as a model of grammatical precision and stylistic elegance fostered the stylistic “improvement” of English in order to make it “equal” to Latin (cf. the varying assessment of English reported e.g. by Rusch 1972). What also played a role in this context is the humanistic rediscovery of the original classical texts and the influence of classical rhetoric in English education (cf. the works of Erasmus and, in English, Wilson; cf. Plett 2004). The lexical elaboration of English, going on since the early Middle Ages but consciously intensified particularly in the 16th and 17th centuries (cf. Nevalainen 1999; Barber 1997), produced stylistic levels within the lexicon, with the Romance elements tending towards more impersonal, abstract, and formal expression versus the native
16 Linguistic Levels: Styles, registers, genres, text types elements’ potentially more down-to-earth, emotive, and colloquial impact. While the lexical division is partly linked to register (e.g. cardio-/cardiac in the medical register vs. heart elsewhere), it can also be exploited for stylistic effects. Hughes (2000) illustrates how (quasi-) synonymous doublets and triplets differ stylistically depending on etymology and time of borrowing. In general one can say that the style of a given piece of writing or passage from at least about the 15th century onwards is in no small part determined by the percentage of Romance and classical vocabulary used (Gordon 1966). The so-called “aureate” style of Lydgate and others in the 15th century is created by a conscious use of elevated Romance lexis, and partly also by complex noun phrases and Latinate syntax. How Shakespeare exploited these lexical distinctions for characterizing persons and situations as well as changing relationships and generally for stylistic effect was shown by Scha¨fer (1973). Besides expanding the means of the language in general, lexical growth had a stylistic aim from the start in being crucially linked to the rhetorical concept of copiousness, i.e. lexical variety through amplification, synonymy, repetition, and paraphrasing.
3.2 Oral vs. literate styles The connection between standardization, writing, texts and the orality-literacy cline was investigated in particular by Biber (1995; cf. also Biber and Finegan 1989, 1992). He examined the development of eight English genres, called “registers” by him (personal essays, medical research articles, science research articles, legal opinions, fiction, personal letters, and dialogue from drama and from fiction) from the 17th to the 20th century based on the three dimensions: (1) Involved vs. Informational Production, (2) Situation-dependent vs. Elaborated Reference, and (3) Non-abstract vs. Abstract Style. These dimensions are characterized by the significant presence or absence of the following features: (1) private verbs, 1P and 2P pronouns, present tense verbs, demonstrative pronouns, be as main verb etc., (2) wh-relative clauses, pied piping, nominalisations, phrasal coordination, (3) conjuncts, passives, past participles, adverbial subordinators (cf. Biber 1988 for a comprehensive feature listing). The features of (1) represent the oral/spoken end of the continuum and those of (2) and (3) the written/literate end. Biber’s results show that all prose registers are clearly non-oral in character already at the beginning of his time frame and that most of them become increasingly more literate during the next 100 or 200 years, especially more informational and elaborated. The 18th century in particular is characterized by very pronounced literate textual realizations. According to Biber (1995: 298), such a development is typical of the early stages of the introduction of writing in a language. As this does not really apply to English in the 17th century, it is rather the combined result of ongoing standardization, in particular the culmination point of its normative phase in
243
244
II Linguistic Levels the 18th century, and perhaps the requirements of the more “public” genres. McIntosh’s (1998) “gentrification” thesis confirms and complements Biber’s results, by tracing a development towards a more literate and elegant prose in the course of the 18th century, a process that is driven by class-consciousness and a feeling of propriety. The resulting “gentrified” style is marked by complex but very orderly sentences, passives, nominalizations, as well as polysyllabic and abstract vocabulary. While Biber’s genres develop largely in parallel until the 18th century, they start to diverge from the 19th century onwards. Medical, scientific, and legal prose consistently proceeded to develop towards even more literate styles up to the 20th century, while the popular, non-expository genres (essays, fiction, drama, letters) gradually and increasingly reversed towards more oral, i.e. involved, non-abstract, and situation-dependent, realizations. In the modern period there is thus a clear stylistic split between the two groups. While the research just summarized deals with a long-range perspective and uses the stylistic dimensions originally produced on the basis of 20th-century material, Biber (2001) reapplied factorial analysis to 18th-century data, thus producing the stylistic dimensions typical for that time. The results point to the fact that there was a more pronounced distinction between spoken and written registers in the past than there is today, with drama but none of the other types being marked as extremely oral. It is also possible that this result is an artefact of the missing spoken dimension, thus making drama stand out in peculiar ways.
3.3 Historical styles: two examples Any literary history will abound with styles, be it the aureate diction of the 15th century, Euphuism in the 16th century or Milton’s Latinate style. Some of these may be of literary interest only, while others are relevant to both literary and linguistic scholars, among which are the clergial or curial style, the so-called “plain style” and the stylistic shift taking place in the 18th century mentioned above. The clergial style, found in the 14th and 15th centuries, derives ultimately from medieval chancellary Latin, the dictaminal arts in general, and more directly from French vernacular models, which were imitated and developed by English writers. It eventually turned into a courtly prose style. It can be found in Chaucer’s prose (Bornstein 1978, with a long list of examples), in writings of the English royal administration and in Caxton’s works (Burnley 1986; Blake 1992). Features of this style include Latinate words and constructions, extensive clausal qualifiers, and long complex sentences, synonymous doublets, anaphoric cohesive devices, epithets, and a preference for the use of the passive. The plain style emerging in the 17th century needs to be looked at in a larger context, as it is partly a reaction against other stylistic models. According to Sprat’s wellknown formulation about the discourse of the Royal Society the motivation for the plain style was to reject all the amplifications, digressions, and swellings of style: to return back to the primitive purity, and shortness, when men deliver’d so many things, almost in an equal number of words. […] a close, naked, natural way of speaking; positive expressions; clear senses; a native easiness: bringing all things as near the Mathematical plainness, as they can: and preferring the language of Artizans, Countrymen, and Merchants before that, of Wits, or Scholars (Sprat 1667: 113).
16 Linguistic Levels: Styles, registers, genres, text types In spite of Sprat’s testimony neither the genesis, the characteristics, nor the type of users of the plain style are entirely clear; partly this may also be due to the fact that the 17th century itself used the term to refer to opposing styles (Adolph 1968: 130). Taavitsainen (2001: 196) characterizes the “house style” of the Philosophical Transactions, the organ of the Royal Society, as marked by first-person narration, subjective point of view, and expressions of low modality. Hu¨llen’s (1989) analysis of some of the descriptive adjectives Sprat uses (e.g. plain, naked, easy) shows that apart from conveying the senses ‘generally understandable’ and ‘unadorned’ they have a number of meanings linking linguistic style to thought, and more generally life styles (e.g. ‘not intellectual, objective, frank, theoretical, contentedness’). For Gordon (1966: 127), this style embodied a rejection of Latinate syntax (Ciceronian and Senecan), of rhetorical figures, and of metaphor and simile, as well as a return to Anglo-Saxon sentence structures – i.e. it is one form of the speech-based prose using ordinary vocabulary and the “grammar of spoken English” (Gordon (1966: 122) that he identified in the 17th century. It is, however, the genteel and polite form produced by educated gentleman, which is what Sprat’s early scientists were; it is not necessarily the speech of merchants or artisans (Gordon 1966: 128; Atkinson 1996: 362–364; Hu¨llen 1989: 84). While Sprat linked the new style to the Royal Society and thus to natural science, it is better seen as a larger and manifold process. A plain style is also found with authors such as John Wilkins, John Webster, and Joseph Glanvill, and in fields such as law, religion, and travel literature, which do not belong to the new science (Hu¨llen 1989: 70), and needs to be seen in the context of a larger development from a rhetoric of persuasion to one of exposition (thus linking up with text types) from the end of the 16th century onwards (Howell 1956: 388). Various authorities, while agreeing on the overall stylistic shift, have emphasized different aspects and times as being decisive for the new style. Fish (1952) stressed the importance of puritan preaching style, shifting the emergence back to as early as 1570, and Warner (1961: 97) saw the likelihood of an influence of the Authorized Version of the Bible, science figures prominently in other explanations. Adolph (1968), placing the shift around 1600, saw the scientific and puritan notions of utilitarianism as decisive for its development. Picking the same time frame, Croll (1921) opted for Francis Bacon’s personal antiCiceronian style as the foundation of the new plain style. Jones (1953) came closest to Sprat by identifying the Restoration period and the emerging sciences and rationalism as the crucial aspects for the new style. Despite the disagreements, some things seem clear: new communicative needs, based on diverse religious practices and on an emerging natural science, but also on newly arising economic and social conditions (e.g. rise of the “middle classes”), led to a new style. This style was less overtly influenced by Latin rhetorical models, broke with “scholastic” (thought)styles (cf. also Taavitsainen 2001), and was potentially more focused on the author. Gordon’s characterization of this style as speech-based, however, conflicts with Biber’s (1995) and Biber and Finegan’s (1989, 1992) findings of the already largely literate character of this period’s prose (cf. above).
4 Approaches to historical texts Kohnen (2008) distinguishes three approaches to the investigation of historical texts: (1) historical text linguistics proper, (2) diachronically oriented text linguistics, and (3) text-oriented historical linguistics.
245
246
II Linguistic Levels The first of these approaches is the study of texts, genres and discourse domains in historical periods of a language, which crucially includes genre inventories. The second approach studies changes in texts, genres and discourse domains across periods, e.g. conventional text structure, or the link between macro-categories and specific text types. The third approach deals with the functions of texts, genres and discourse domains in language change.
4.1 Historical text linguistics: genre histories Go¨rlach (1992, 2001b, 2002, 2004) takes an external approach to the question of genres, although he uses the term “text type”. His suggestion is to collect genre names existing at a given time or throughout the history and to conduct a componential analysis on the lexical field found. Go¨rlach (2001b, 2004) gives the following 24 parameters (thought of as equivalent to semantic markers) for distinguishing text types, to which I add in brackets his characterization of the genre “contract”: field (law), intention (binding), act/ action connected to text (none), accompaniment by music or visual material (no), conglomerate (no), composite (yes), boundness (free), cohesion/coherence (yes), original (yes), nativeness (yes), general/regional (standard), time (present-day), medium (written), style (formal), form (prose), formulaicness (yes), orientation (content-oriented), specialization (technical), truth (non-fictional), spontaneity (no: revised), publicness (yes), length (n.a.), and official (yes). Go¨rlach (2004) provides two alphabetical lists of English genre terms based on entries in the Shorter Oxford English Dictionary (with dates) and on the Thesaurus of Old English. Examples of genre terms from Go¨rlach’s list are eulogy, handbook, invoice, joke, lecture, parable, report, and small talk. He also illustrates his approach with an exemplary componential analysis of genre terms beginning with , from which the contract example above was taken. The presence or absence of a genre term at a given time, extractable from Go¨rlach’s list, and its understanding by contemporaries themselves highlight the cultural determinacy of the concept genre. Genres and text types may remain fairly stable over a long time (e.g. the sonnet), but equally they may be falling into disuse (e.g. those marked by a dagger in Go¨rlach’s list, or the telegram at present) or be newly emerging (e.g. the essay in the 16th century) (cf. Go¨rlach 2002). Some types evolve only within (newly established) media, registers, or publication types, such as dedications within books (15th century), and a whole variety within newspapers (e.g. editorial 1830, obituary 1828, weather forecast 1883; Go¨rlach 2004; see Fries Chapter 67), though sometimes taking existing genres as models (e.g. early dedications and news reports being connected to letters).
4.2 Diachronically oriented text linguistics Conventional and changing textual structures have been investigated for the genre of letters, for example. Richardson (1984) describes Chancery and other letters, based on Latin style, to be structured as follows: (1) address, (2) salutation, (3) notification,
16 Linguistic Levels: Styles, registers, genres, text types (4) (5) (6) (7) (8) (9)
exposition, disposition/ disjunction, final clause, valediction, attestation and date.
Davis (1965), in contrast, argues for a French model to be followed in correspondence, consisting of (1) (2) (3) (4)
address, commendation formula, health formula at the beginning of the letter, and closing formula.
Nevalainen (2001) explores the use of these formulae in the letters of the Johnson merchant family written 1542–1552. She finds that the Johnson letters follow six of the nine medieval letter-writing conventions, namely date (and place), salutation, address, “health” formula, notification, and (a form of) valediction, but they do this to varying degrees, with great individual variation and with additions and modifications. Okulska (2006) deals with structural letter elements, but also with topic development and information structure in the sub-genre of the diplomatic narrative report letter, from the 15th to the 18th century. With regard to the latter aspect, the letters represent discursive hybrids combining narrative (past orientation, topic-based) and reporting (present tense, person-centered) elements, the latter also frequently triggering evaluative-argumentative comments in the early, but not the later, part of the period. As to the presentation of information in these letters, there is a shift from inductive topic-delayed to deductive topic-first thesis presentation from the 16th century onwards, which puts the focus more clearly on the information itself than on the persons interacting. Another approach with regard to text types is to follow the diachronic development of the prototypical text types, such as narrative or exposition, or, as there are hardly pure types, of genres which are typically dominated by one of these, e.g. works on language as representing the instructive type. This has not been done for any of the types on a broad scale so far, but smaller-scale treatments exist. Taavitsainen (2004), for example, deals with genres of secular instruction from Old English to Early Modern English, in particular with how they are made more pleasant and entertaining for the readers. This means the inclusion of other than purely instructive features, such as the use of verse form, narrative passages, dialogues, typical involvement features, and a conventional fictionalized frame. Taavitsainen found both considerable variation within the text category and interesting diachronic continuities, as well as a “circular movement” of features moving from learned texts to more popular texts, thus again leaving space for innovation at the learned end of the continuum. With regard to narrative discourse, for example, Wa˚rvik (1990a and 1990b) describes the history of grounding markers in English narratives. She finds that English has undergone a change from a foreground-signalling language in Old English (by use of the more or less obligatory marker þa) to a background-signalling language in modern English
247
248
II Linguistic Levels (by means of, e.g., subordination, non-finite forms). Whereas marking in Old English is fairly specific and, in a certain sense, mono-dimensional, the modern system is more fuzzy, being characterized by a variety of non-obligatory, partly stylistic options. A further avenue of research investigates the aspect that certain text categories or genres seem to prefer certain text-level features rather than others. One good example for this is discourse deixis, which has been investigated by Fries (1993, 1994), Claridge (2001), and Kilpio¨ (1997). Fries (1993, 1994) found deictic elements marking text location very or most commonly in the text categories instruction (religious and secular), non-imaginative narration, and exposition in Old English and Early Modern English. The markers afore and the said very clearly dominate in statutory texts. Statutory and similar texts also figure prominently in Kilpio¨’s (1997) diachronic investigation, which concentrates on participial adjectives of the type (a) forementioned in the entire Helsinki Corpus (HC) (thus using the Helsinki Corpus text-labelling terminology). While these discourse-deictic elements are rare in Old English and early Middle English, they become more frequent from 1350 onwards with genres like official letters, documents, and law heading the frequency lists (joined by history and science in Early Modern English). In contrast, Old English law does not use such features. Of interest are also the kinds of genres which make do without these discourse-deictic elements, namely Middle English and early Modern English rule, Bible, homily, drama, comedy, and those that use them very rarely, namely sermons, fiction, and education. Sermons and religious treatises also show the lowest instance of discourse-deictic terms in Claridge’s (2001) study, while the legal register, represented by many different genres in the corpus, has the highest number. Early Modern English scientific genres show a high number of items indicating present location and forward-indicating items, pointing to the importance of commenting on the ongoing discourse procedure of the author. The study of connective devices is another case in point for text-level feature preferences insofar as it illuminates the kinds of explicit cohesive relationships typical for certain genres. Kohnen (2007) investigated what he termed “connective profiles”, i.e. the overall use of the whole range of coordinating and subordinating conjunctions/complementizers. Comparing 17th-century sermons and statutes he found quite distinct profiles: a fairly high and heterogeneous number of connectives in sermons and a much lower, less varied use in statutes. Partly this seems to be due to a stylistic preference for conjunctionless non-finite subordination in statutory language. But it is also based on the communicative functions common of statutes. Rare but and non-existent for are expendable because contrastive viewpoints and the provision of justifications are hardly found in these texts, whereas moderately common and is useful for enumerations. All subordinators are less common in statutes than in sermons, except for manner/comparison markers, which is due to the legal formulaic use of introductory whereas/as clauses. A notable frequency is furthermore only exhibited by nominal clause connectives (especially that) and by conditional clauses. Kohnen also looked at the diachronic development of sermons over the 15th, 16th, 17th, and late 20th century, and sums up the general pattern as a decrease in oral features and simultaneous increase in literate features in Early Modern English, followed by the reverse development in the 20th century.
16 Linguistic Levels: Styles, registers, genres, text types
4.3 The role of texts in language change Kohnen (2001) sees genres (“text types” in his terminology) as catalysts in language change, i.e. they facilitate change and are responsible for the spreading of a construction, his example being the adverbial first participle construction. He showed how this construction spread across text types over a period of two centuries, its textual frequency first increasing in religious treatises around 1340, then about 1390 in both homilies/sermons and petitions/statutes, by roughly 1470 in chronicles, and finally by around 1520 in narrative prose and private letters. In each instance the construction fulfils an important functional requirement of the text type, e.g. introducing explanatory passages in Biblical exegesis (treatises), providing vivid descriptions in sermons, petitions, and narrative prose, or encoding formulaic speech acts in the introductory sections of letters. According to Kohnen, the particular chronological adoption of the feature by the genres is due to the following three aspects. Early or well-established vernacular types, such as religious treatises and sermons, make use of the participle construction earlier than types which are found only later as vernacular forms, e.g. letters. So do genres close to Latin and/or French traditions (both of which had the construction in question) and those that are more formal in character, such as religious prose and petitions/ statutes. Lastly, the linguistic features adopted by genres with considerable social relevance and a prestigious status, i.e. religious treatises, statutes, and documents, are more likely to spread to other text types later than features adopted first by less prestigious texts.
5 Summary The above presentation has somewhat artificially separated things that intimately belong together. A style is found in a text which belongs both to a register and to a genre. Ideally, these aspects should be treated together then, but this would increase the complexity to such an extent that larger-scale and diachronic investigations would be difficult. Thus, particular research efforts tend to concentrate on only one of these aspects, even if this necessarily means simplifying. Another complicating factor, also visible in the extant research, is the variable usage of the basic terminology – what is one researcher’s style may be another’s register, which makes comparisons across works difficult. Furthermore, despite Kohnen (2008) quoted above, there is as yet no historical or diachronic text linguistics as an established field. Researchers in this field may at the moment tend to place themselves vaguely in the paradigm of historical pragmatics. In contrast to the theoretical difficulties, actually researching historical texts may be getting increasingly easier. Many of the researchers quoted above approach the field with the help of corpus linguistics. A range of historical corpora (Kyto¨, Chapter 96, Volume 2,) is available by now, all of which use some kind of register, text type or genre coding system. A Representative Corpus of Historical English Registers (ARCHER) (Biber and Finegan 1990–93/2002/2007/2010) has been compiled and is being extended by Biber and various affiliates. The Helsinki Corpus (HC) (Rissanen et al. 1991) contains texts from all the registers mentioned. As more general corpora often do not contain individual registers in sufficient numbers, single-register corpora are a particular asset. The Zurich English Newspaper Corpus (ZEN) (Fries et al.
249
250
II Linguistic Levels 2004), the Lampeter Corpus (pamphlets) (Claridge et al. 1999) and the Lancaster Newsbook Corpus (McEnery and Hardie 2001–07) can be used together to investigate the origins of the press register. Similarly, the Corpus of Early English Medical Writing (CEEM) (Taavitsainen et al. 1995–), compiled at Helsinki University, charts the development of a scientific writing based on the medical prototype. The Corpus of Early English Correspondence Sampler (CEECS) (Kera¨nen et al. 1998) could be called a single-genre corpus. Not all of the above-mentioned corpora contain complete texts, however, which would be of especial importance for the genre or text-type approach.
6 References Adolph, Robert. 1968. The Rise of Modern Prose Style. Cambridge, MA: MIT Press. Archer, Dawn. 2005. Questions and Answers in the English Courtroom (1640–1760). Amsterdam/ Philadelphia: John Benjamins. Atkinson, Dwight. 1996. The Philosophical Transactions of the Royal Society of London, 1675– 1975. Language in Society 25: 333–371. Barber, Charles. 1997. Early Modern English. Edinburgh: Edinburgh University Press. Biber, Douglas. 1988. Variation Across Speech and Writing. Cambridge: Cambridge University Press. Biber, Douglas. 1989. A typology of English texts. Linguistics 27(1): 3–43. Biber, Douglas. 1995. Dimensions of Register Variation. Cambridge: Cambridge University Press. Biber, Douglas. 2001. Dimensions of variation among 18th-century registers. In: Diller and Go¨rlach (eds.), 89–109. Biber, Douglas, and Edward Finegan. 1989. Drift and the evolution of English style: A history of three genres. Language 65: 487–517. Biber, Douglas and Edward Finegan. 1990–93/2002/2007/2010. A Representative Corpus of Historical English Registers (ARCHER). Version 3.1. Consortium of fourteen universities, see http:// www.llc.manchester.ac.uk/research/projects/archer/ Biber, Douglas, and Edward Finegan. 1992. The linguistic evolution of five written and speechbased English genres from the 17th to the 20th centuries. In: Rissanen et al. (eds.), 688–704. Blake, Norman. 1992. The literary language. In: Norman Blake (ed.), Cambridge History of the English Language, Vol. II. 1066–1476, 500–541. Cambridge: Cambridge University Press. Bornstein, Diane. 1978. Chaucer’s Tale of Melibee as an example of “style clergial”. The Chaucer Review 12: 236–254. Burnley, J. David. 1986. Curial prose in England. Speculum 61(3): 593–614. Chafe, Wallace. 1982. Integration and involvement in speaking, writing, and oral literature. In: Deborah Tannen (ed.), Spoken and Written Language: Exploring Orality and Literacy, 35–53. Norwood, NJ: Ablex. Claridge, Claudia, Josef Schmied, and Rainer Siemund. 1999. The Lampeter Corpus of Early Modern English Tracts. In: ICAME Collection of English Language Corpora (CD-ROM), 2nd edn., Knut Hofland, Anne Lindebjerg, and Jørn Thunestvedt (eds.), The HIT Centre, University of Bergen, Norway For manual see http://khnt.hit.uib.no/icame/manuals/LAMPETER/ LAMPHOME.HTM Claridge, Claudia. 2001. Structuring text: Discourse deixis in Early Modern English Texts. Journal of English Linguistics 29/1: 55–71. Croll, Morris W. 1921. “Attic Prose” in the seventeenth century. Studies in Philology 18: 79–128. Davis, Norman. 1965. The Litera Troili and English letters. Review of English Studies, New Series 16: 233–244. Diller, Hans-Ju¨rgen. 1998. Stylistics: linguistic and textual. European Journal of English Studies 2(2): 155–174.
16 Linguistic Levels: Styles, registers, genres, text types Diller, Hans-Ju¨rgen. 2001. Genre in linguistic and related discourses. In: Diller and Go¨rlach (eds.), 3–43. Diller, Hans-Ju¨rgen and Manfred Go¨rlach (eds.). 2001. Towards a History of English as a History of Genres. Heidelberg: Winter. Fish, Harold. 1952. The Puritans and the reform of prose-style. A Journal of English Literary History 19: 229–248. Fisher, John. 1996. The Emergence of Standard English. Lexington: University Press of Kentucky. Fries, Udo. 1993. Towards a description of text deixis in Old English. In: Klaus R. Grinda and Claus-Dieter Wetzel (eds.), Anglo-Saxonica: Beitra¨ge zur Vor- und Fru¨hgeschichte der englischen Sprache und zur altenglischen Literatur, 527–540. Munich: Fink. Fries, Udo. 1994. Text deixis in Early Modern English. In: Dieter Kastovsky (ed.), Studies in Early Modern English, 111–128. Berlin/New York: Mouton de Gruyter. Fries, Udo, Hans Martin Lehmann et al. 2004. Zurich English Newspaper Corpus (ZEN). Version 1.0. Zu¨rich: University of Zu¨rich. http://es-zen.unizh.ch http://www.helsinki.fi/varieng/CoRD/ corpora/ZEN/index.html (see also: http://www.helsinki.fi/varieng/CoRD/corpora/ZEN/index. html). Gordon, Ian A. 1966. The Movement of English Prose. London: Longman. Go¨rlach, Manfred. 1992. Text-types and language history: the cookery recipe. In: Rissanen et al. (eds.), 736–761. Go¨rlach, Manfred. 1999a. English in Nineteenth-century England. Cambridge: Cambridge University Press. Go¨rlach, Manfred. 1999b. Regional and social variation. In: Lass (ed.), 459–538. Go¨rlach, Manfred. 2001a. Eighteenth-century English. Heidelberg: Winter. Go¨rlach, Manfred. 2001b. A history of text types: A componential analysis. In: Diller and Go¨rlach, 47–88. Go¨rlach, Manfred. 2002. What’s in a name? Terms designating text types and the history of English. In: Andreas Fischer, Gunnel Tottie, and Hans-Martin Lehmann (eds.), Text Types and Corpora. Studies in Honour of Udo Fries, 17–28. Tu¨bingen: Narr. Go¨rlach, Manfred. 2004. Text Types and the History of English. Berlin/New York: Mouton de Gruyter. Grund, Peter. 2007. From tongue to text: The transmission of the Salem Witchcraft Examination Records. American Speech 82: 119–150. Halliday, M. A. K. 1988. On the language of physical science. In: Mohsen Ghadessy (ed.), Registers of Written English, 162–178. London: Pinter. Haugen, Einar. 1972. Dialect, language, nation. In: J. B. Pride and Janet Holmes (eds.), Sociolinguistics, 97–111. Harmondsworth: Penguin. Hiltunen, Risto. 1990. Chapters on Legal English. Aspects Past and Present of the Language of the Law. Helsinki: Suomalainen Tiedeakatemia. Howell, W. S. 1956. Logic and Rhetoric in England, 1500–1700. New York: Russell & Russell. Hughes, Geoffrey. 2000. A History of English Words. Oxford: Blackwell. Hu¨llen, Werner. 1989. “Their Manner of Discourse”: Nachdenken u¨ber Sprache im Umkreis der Royal Society. Tu¨bingen: Narr. Hunter, Michael. 1981. Science and Society in Restoration England. Cambridge: Cambridge University Press. Jones, Richard Foster. 1953. The Triumph of the English Language. Stanford: Stanford University Press. Kera¨nen, Jukka, Minna Nevala, Terttu Nevalainen, Arja Nurmi, Minna Palander-Collin and Helena Raumolin-Brunberg. 1998. Corpus of Early English Correspondence Sampler (CEECS). In: ICAME Collection of English Language Corpora (CD-ROM), 2nd edn., Knut Hofland, Anne Lindebjerg, and Jørn Thunestvedt (eds.), The HIT Centre, University of Bergen, Norway. For manual see http://khnt.hit.uib.no/icame/manuals/ceecs/INDEX.HTM (See also: http://www.helsinki.fi/varieng/domains/CEEC.html)
251
252
II Linguistic Levels Kilpio¨, Matti. 1997. Participial adjectives with anaphoric reference of the type the said, the (a) forementioned from Old to Early Modern English: The evidence of the Helsinki Corpus. In: Terttu Nevalainen and Leena Kahlas-Tarkka (eds.), To Explain the Present: Studies in the Changing English Language in Honour of Matti Rissanen, 77–100. Helsinki: Socie´te´ Ne´ophilologique. Kohnen, Thomas. 2001. Text types as catalysts for language change: The example of the adverbial first participle construction. In: Diller and Go¨rlach (eds.), 111–124. Kohnen, Thomas. 2007. “Connective profiles” in the history of English texts. In: Ursula Lenker and Anneli Meurman-Solin (eds.), Connectives in the History of English, 289–308. Amsterdam/Philadelphia: John Benjamins. Kohnen, Thomas. 2008. Historical text linguistics: Investigating language change in texts and genres. Paper delivered at ICEHL 15, Munich. Lass, Roger (ed.). 1999. Cambridge History of the English Language, Vol. III. 1476–1776. Cambridge: Cambridge University Press. Liebermann, Felix. 1903–16. Die Gesetze der Angelsachsen. 3 vols. Halle an der Saale: Max Niemeyer. Longacre, Robert E. 1996. The Grammar of Discourse. 2nd edn. New York: Plenum Press. McEnery, Tony and Andrew Hardie. 2001–07. Lancaster Newsbooks Corpus. UCREL and Linguistics and English Language, University of Lancaster. http://www.ling.lancs.ac.uk/activities/ 695/. Available through the Oxford Text Archive: http://ota.ahds.ac.uk/headers/2531.xml McIntosh, Carey. 1998. The Evolution of English Prose, 1700–1800: Style, Politeness, and Print Culture. Cambridge: Cambridge University Press. Mellinkoff, David. 1963. The Language of the Law. Boston: Little, Brown. Moessner, Lilo (ed.). 2001. Special Issue of European Journal of English Studies 5(2). Nevalainen, Terttu. 1999. Early Modern English lexis and semantics. In: Lass (ed.), 332–458. Nevalainen, Terttu. 2001. Continental conventions in early English correspondence. In: Diller and Go¨rlach (eds.), 203–224. Okulska, Urszula. 2006. Textual strategies in the diplomatic correspondence of the Middle and Early Modern English periods: The narrative report letter as a genre. In: Marina Dossena and Susan Fitzmaurice (eds.), Business and Official Correspondence: Historical Investigations, 47–76. Bern: Lang. Plett, Heinrich F. 2004. Rhetoric and Renaissance Culture. Berlin/New York: Mouton de Gruyter. Richardson, Malcolm. 1984. The dictamen and its influence on fifteenth-century English prose. Rhetorica 2: 207–226. Rissanen, Matti. 1999. Syntax. In: Lass (ed.), 187–331. Rissanen, Matti, Merja Kyto¨, Leena Kahlas-Tarkka, Matti Kilpio¨, Saara Nevanlinna, Irma Taavitsainen, Terttu Nevalainen, and Helena Raumolin-Brunberg. 1991. The Helsinki Corpus of English Texts (HC). In: ICAME Collection of English Language Corpora (CD-ROM), 2nd edn., Knut Hofland, Anne Lindebjerg, and Jørn Thunestvedt (eds.), The HIT Centre, University of Bergen, Norway. For manual, see http://khnt.hit.uib.no/icame/manuals/hc/index.htm Rissanen, Matti, Ossi Ihalainen, Terttu Nevalainen, and Irma Taavitsainen (eds.). 1992. History of Englishes: New Methods and Interpretations in Historical Linguistics. Berlin/New York: Mouton de Gruyter. Rusch, Ju¨rg. 1972. Die Vorstellung vom Goldenen Zeitalter der englischen Sprache im 16., 17., und 18. Jahrhundert. Bern: Francke. Samuels, M. L. 1963. Some applications of Middle English dialectology. English Studies: A Journal of English Language and Literature 44: 81–94. Scha¨fer, Ju¨rgen. 1973. Shakespeares Stil: germanisches und romanisches Vokabular. Frankfurt: Athena¨um. Skaffari, Janne, Matti Peikola, Ruth Carroll, Risto Hiltunen, and Brita Wa˚rvik (eds.). 2005. Opening Windows on Texts and Discourses of the Past. Amsterdam/Philadelphia: John Benjamins.
16 Linguistic Levels: Styles, registers, genres, text types Sprat, Thomas. 1667. The history of the Royal-Society of London for the improving of natural knowledge by Tho. Sprat. London: Printed by T. R. for J. Martyn …, and J. Allestry … (accessible via EEBO: http://eebo.chadwyck.com/home). Taavitsainen, Irma. 2001. Language history and the scientific register. In: Diller and Go¨rlach (eds.), 185–202. Taavitsainen, Irma. 2002. Historical discourse analysis: Scientific language and changing thoughtstyles. In: Teresa Fanego, Bele´n Me´ndez-Naya, and Elena Seoane (eds.), Sounds, Words, Texts and Change, 201–226. Amsterdam/Philadelphia: John Benjamins. Taavitsainen, Irma. 2004. Genres of secular instruction: A linguistic history of useful entertainment. Miscela´nea: A Journal of English and American Studies 29: 75–94. Taavitsainen, Irma. 2005. Genres and the appropriation of science: Loci communes in English in the late medieval and early modern period. In: Skaffari et al. (eds.), 179–196. Taavitsainen, Irma, Pa¨ivi Pahta, Martti Ma¨kinen, Turo Hiltunen, Ville Marttila, Maura Ratia, Carla Suhr, and Jukka Tyrkko¨. forthc. Corpus of Early English Medical Writing (CEEM). University of Helsinki. http://www.helsinki.fi/varieng/CoRD/corpora/CEEM/index.html Warner, Alan. 1961. A Short Guide to English Style. Oxford: Oxford University Press. Wa˚rvik, Brita. 1990a. On grounding in English narratives: A diachronic perspective. In: Sylvia Adamson, Vivien Law, Nigel Vincent, and Susan Wright (eds.), Papers from the 5th International Conference on English Historical Linguistics, Cambridge, 6–9 April 1987, 559–575. Amsterdam/ Philadelphia: John Benjamins. Wa˚rvik, Brita. 1990b. On the history of grounding markers in English narrative: Style or typology? In: Henning Andersen and Konrad Koerner (eds.), Historical Linguistics 1987, 531–542. Amsterdam/Philadelphia: John Benjamins. Werlich, Egon. 1983. Text Grammar of English. 2nd edn. Heidelberg: Quelle & Meyer. Wright, Laura. 1992. Macaronic writing in a London archive, 1380–1480. In: Rissanen et al. (eds.), 762–779.
Claudia Claridge, Duisburg-Essen (Germany)
253
III Old English 17 Old English: Phonology 1 2 3 4 5 6
Introduction Terminology, evidence, methods Synchrony Phonological–orthographic correspondences Diachrony References
Abstract The investigation of Old English phonology has been incessant over the decades and carried out from the vantage point of many different theoretical perspectives, but it remains a remarkable fact of our neogrammarian legacy that fundamental aspects of their Old English analyses have weathered the changing theoretical winds particularly well. This high degree of consensus allows us to present a very broad, relatively uncontroversial overview of many fundamental aspects of OE phonology. At the same time, of course, problematic areas remain, and some controversies of perennial interest are indicated.
1 Introduction Old English (OE) phonology has been the subject of scientific investigation for well over 100 years. At the end of the 19th century, an informal group of linguists and philologers based in Leipzig and known as the “neogrammarians” (Ger. Junggrammatiker) constituted the dominant force in linguistic science. Their primary interest was the study of language change – especially sound change in light of the budding science of phonetics (see Sievers 1901) – and a significant amount of their scholarly attention was directed at the investigation of the earliest stages of the Germanic languages. Much of this work was codified in grammars and historical handbooks, which – in keeping with the neogrammarian emphasis – focused primarily on phonological reconstruction and sound change, as well as inflectional morphology. As an early Germanic dialect with a relatively long documented history that included significant literary works such as Beowulf, Old English was of natural interest to the neogrammarians. Eduard Sievers – perhaps the most accomplished and renowned neogrammarian – published his Angelsa¨chsische Grammatik (‘Old English Grammar’) in 1882, and other grammars and classic works strongly influenced by the neogrammarian perspective followed, such as Sweet (1888), Bu¨lbring (1902), Wright and Wright (1925), Campbell (1959), and Brunner (1965 – the last revision of Sievers’s grammar). Luick (1964a [1914–21], 1964b [1929–40]) is of special note in its copious treatment of the sound changes through the entire history of English from its Proto-Indo-European origins to the modern stages. (Luick 1964a [1914–1921]: 94–320 treats phonology and sound changes up to the end of the Old English period.) Although there were many areas of disagreement, for the main part our classic grammars and handbooks reflect a kind of rough neogrammarian consensus Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 255–272
256
III Old English achieved by the early 20th century after decades of intensive research. These works, and especially the much more comprehensive studies on which they are built, remain invaluable research tools to the present day, which in many respects have not been superseded.
2 Terminology, evidence, methods Before presenting our overview of OE phonology, a brief discussion of terminology is in order. In fact, the designation “Old English” is a very broad cover term that abstracts away from the dialectal and chronological realities. The OE period is traditionally set at 700–1100 CE, and four major dialect groups are recognized: Northumbrian, Mercian, West Saxon, and Kentish (where Northumbrian and Mercian together form Anglian). As such, it is sometimes not particularly meaningful to speak of OE phonology without qualification, given the inherent dialectal diversity and the significant changes that occurred over the four hundred year time span. In fact, most traditional descriptions focus on the best-documented dialect group, West Saxon, in which one variant became a kind of standard language around the end of the 10th century. This classical Old English, best represented in the works of Ælfric, is the default reference point for our treatment. It is also worth noting here that there is no direct line of descent from this classical Old English to any of the modern standard varieties, which are based primarily on the Mercian dialect group (see Hogg 1992a: 83–84 and Hogg 2006, Lass 1994: 1–5; see also Sauer and Waxenberger, Chapter 22; Kornexl, Chapter 24). We also briefly consider the types of evidence used in the reconstruction of OE phonology (see Lass 1992: 27–32 for a good discussion of the role of various kinds of evidence). It is an unfortunate fact of the Old and Middle English periods that we have no direct descriptions of the spoken form of any dialect, since such descriptions of English only begin in the 16th century. Accordingly, the most direct evidence available for the OE period is primarily of three types: spelling, poetic metrical conventions, and borrowing. In the case of writing, although it is an exaggeration to state that the “scribes wrote as they spoke”, spelling during the OE period was more or less phonologically based, so generally there was a fairly close match between the respective phonological and written forms. This fact, along with the use of the Roman alphabet, means that the first level of analysis – that is, a very broad phonological rendering – is relatively straightforward at the segmental level, although of course not entirely without controversy. At the same time, orthographic systems are inevitably deficient in various ways, and some properties, such as vowel length, were not indicated by the OE scribes. Vowel length, however, can be reconstructed on the basis of the metrical principles derived from such poetic works as Beowulf. The meter of this type of poetry is based on a crucial distinction between light and heavy syllables, which depends in part on the contrast between a short and a long vowel in an open syllable. For example, the first syllables in the words cwe˘.ne ‘woman’ and de¯.man ‘to judge, deem’ function metrically as light and heavy, respectively (see Section 3.6 below). Further corroboration of a length contrast is found in the borrowing of words from Latin into Old English, which although quite limited in scope allows for some cross-referencing in light of our more detailed knowledge of Latin phonology. For example, the borrowing of Lat. nō na as OE nō n ‘noon’, beyond suggesting a phonetic similarity in the quality of
17 Old English: Phonology
257
Lat. and OE ō , dovetails more generally with the reconstruction of a vowel length contrast for both languages. Most importantly, though, the task of reconstructing the synchrony and diachrony of OE phonology is governed by the general principles and methods of linguistic reconstruction (Fox 1995). Fundamental principles here are uniformity (reconstructed systems must be compatible with our knowledge of present-day systems), plausibility (all assumed sound changes must be well motivated), and regularity (generalized regular sound change is assumed, all things being equal).
3 Synchrony 3.1 Vowels In the neogrammarian tradition, a system consisting of seven vowels and contrastive length is reconstructed for Old English (Table 17.1). An additional vowel, [ø] (usually written ), occurred in very early West Saxon, but by the classical OE period it had undergone derounding and merger with e ([ø] survived longer in Mercian and Northumbrian); for example, twoelf ~ twelf ‘twelve’, fo¯et ~ fe¯t ‘feet’ (Hogg 1992b: 124–126). From a typological perspective, the posited Old English vowel system is not particularly unusual, finding a close modern parallel, for example, in Finnish (Maddieson 1984: 275). Table 17.1: Old English vowel system
High Mid Low Long ı¯s ‘ice’ fe¯dan ‘to feed’ dæ¯d ‘deed’ hy¯dan ‘to hide’
Unrounded Front
Rounded
i e æ
y
Unrounded Back
Rounded u o
a Short fı˘sc ‘fish’ le˘þer ‘leather’ ɡlæ̆ d ‘ɡlad’ dy̆ ppan ‘to dip’
Long hlu¯d ‘loud’ bro¯þor ‘brother’ sa¯da ‘snare’
Short hu˘nd ‘hound, dog’ bo˘ɡa ‘bow’ sa˘dol ‘saddle’
Although traditionally duration is considered the primary property distinguishing the Old English short and long vowel pairs, it is common – especially in textbooks on the history of English – to find confident statements about tense/lax distinctions. For example, Pyles and Algeo (1993: 103) state that the short vowels “were approximately [ɛ], [ɪ], [ɔ], and [ʊ] respectively, as in net, nit, nought, and nut”. A more systematic and comprehensive type of challenge to the traditional position was initiated by Stockwell (1952), who analyzed the contrast between the vowel pairs in terms of simple vs. complex (diphthongal) nuclei rather than in terms of duration (Lass and Anderson 1975: 201–205 remains a good discussion of the general issues; see also Hogg 1992a: 85–86 for a discussion of possible special properties of the low vowels). Nevertheless, although the reconstruction of some type of tense/lax contrast is intuitively appealing from the perspective of most present-day standard varieties of English, the phonetic and phonological interplay of vowel length
258
III Old English and quality is an extremely complex area (see, for example, Rosner and Pickering 1994). While it is not uncommon for quality differences (usually described in terms of tenseness or peripherality) to accompany a vowel length contrast, modern languages such as Finnish attest to the fact that salient quality differences are not necessary concomitants of vowel length. In fact, a comparison of Old and Middle English sound changes supports the traditional reconstruction’s focus on the durational – as opposed to qualitative – basis of the OE vowel pairs. In late Old English, in a change known as Homorganic Cluster Lengthening (see also Section 5.2 below), an original short vowel lengthened before clusters such as -ld, -nd, and -mb. In this lengthening, no vowel quality change is evident, as in fe¯ld (< fe˘ld) ‘field’, gru¯nd (< gru˘nd) ‘ground’, and clı¯mban (< clı˘mban) ‘to climb’, thus implying the qualitative matching of the short/long pairs. By contrast, in Middle English, when another set of vowel lengthenings occurred (Open Syllable Lengthening), quality differences are apparent; for example, [e˘] > open [ɛ̄ ] (not [e¯]), as in bre˘ken > bre˛̄ ken ‘to break’); [u˘] > [o¯], du˘res > dō res ‘doors’; and [ı˘] > [e¯], wı˘kes > wē kes ‘weeks’. ˙ ˙ This later Middle English treatment implies significant quality differences between the short and long counterparts, and justifies the assumption that the vowel pairs were no longer paired strictly in terms of duration. In sum, although by early Middle English an increasing qualitative differentiation between the vowel pairs is evident, for Old English the default assumption is a contrast built primarily on duration. Phonologically, the difference can be represented in terms of mono- versus bimoraicity; that is, there are seven vowels, each participating in the suprasegmental length contrast (where μ represents a mora), as in (1): (1)
Vowel length contrast μ
μ
e
μ e
3.2 Diphthongs Classical Old English had only two diphthongs, usually written and , as in de¯op ‘deep’ and de¯aþ ‘death’. These falling or off-gliding diphthongs, like the vowels, carried a length contrast, as confirmed, for example, by their historical development and participation in the syllable weight conventions of OE poetic meter. The length difference was a property of the overall diphthong, not of the individual segments; that is, e͝o vs. e͞o, not e˘o vs. e¯o or eo˘ vs. eo¯ (following convention we continue to use e˘o and e¯o below). The quality of the off-glide is difficult to determine, but the practice of writing and in older manuscripts suggests an original high rounded [w]-like element, which was subject to further assimilatory and reductive changes over time (Lass 1994: 50). In fact, although not represented orthographically, the primary difference between the two diphthongs lay in the quality of the initial portion; that is, began with a mid e-type vowel and with a more open æ-type. This qualitative difference is suggested by the orthographic variation in older manuscripts in which
17 Old English: Phonology alternates with , and with both and . It is also suggested by the fact that some instances of eo and ea derive historically from earlier e and a, respectively, through sound changes. For example, in a change traditionally known as “breaking”, diphthongization of the vowel occurs when it is followed by specific consonants (l, r, h), as in e˘orþe (e˘o < e˘) ‘earth’ and eahta (e˘a < a) ‘eight’. Accordingly, the standardized use of (as opposed to a more phonologically accurate rendering, or ) can be considered an artifact of a scribal preference for avoiding the repetition of certain graph sequences. Most of the confident description presented in the previous paragraph can be gleaned from our earliest grammars and reference works (for example, Sievers 1898: 14; 1901: 194–195, 293; and Luick 1964a [1914–21]: 138). However, one seemingly innocuous aspect of the traditional view – namely, the assumption of a length contrast – has proved very troubling to succeeding generations of linguists and resulted in a massive amount of scholarly attention apparently disproportionate to the importance of two humble diphthongs destined to disappear without a trace in Middle English (Hogg 1992a: 104). In fact, however, the scholarly effort is justified in that the reconstruction of diphthongal length raises important theoretical and typological issues, given the apparent rarity – or absence according to White (2004) – of such a contrast in the languages of the world. From a uniformitarian perspective, the failure to find a parallel in any modern language would strongly suggest the inappropriateness of the reconstruction for an obsolescent language. Skepticism relating to phonological plausibility has been at the root of numerous attempts to revise the traditional analysis, beginning with Daunt (1939) whose interpretation is developed on the assumption that OE scribes adopted the Irish scribal practice of using vowels as diacritics to indicate specific qualities of an immediately following consonant; that is, the primary function of the second part of the short digraph was not to indicate diphthongal qualities, but rather the backness of the following consonant. In other words, the OE phonological system did not contain short diphthongs at all, and the digraphs and , when “short”, were simply an orthographic convention (the existence of the long diphthongs is not questioned). Intense debate on this topic continued over the decades – especially under the rubric of American structuralism; see, for example, Stockwell and Barritt 1955 and Hockett 1959 – but no consensus was ever reached. (For a good overview of the issues and literature, see Hogg 1992b: 16–24.) Part of the difficulty in reaching a consensus is that the plausibility issues are not as easily resolved as one might naively expect. The empirical waters are muddy in part because diphthongs are theory-dependent entities. For example, length contrasts involving apparent diphthongal pairs such as e˘w vs. e¯w are relatively common in languages. However, if for example e¯w is analyzed as a sequence of V̅ + C – that is, not as a “true” diphthong involving a complex nucleus – then the existence of such pairs becomes irrelevant to the OE situation. Furthermore, in cases where diphthongal length is assumed, it is often argued that the contrast rests on an opposition of long vs. overlong, rather than on short vs. long as assumed for Old English. For example, in a language such as standard Thai (which has been analyzed with a diphthongal length contrast) the short/long diphthongs are indeed phonetically longer than the simple short/long vowel pairs. At the same time, however, both vowels and diphthongs pattern in the same way in terms of the length contrast – that is, there is close to a 1:2 durational difference between short and long pairs regardless of whether simple vowels or diphthongs
259
260
III Old English are involved (Noss 1964: 15). Thus, given that a diphthongal length contrast is possible in principle, for a language such as Old English – with its strict syllable weight system – there would seem to be nothing implausible about pairs of diphthongs conforming phonologically to the short/long system, as opposed to building some other independent sub-system such as long vs. overlong. Although some scholars (for example, White 2004) have attempted to reject the possible existence of short diphthongs on theoretical grounds, the arguments are difficult to sustain. If the primary difference between a diphthong and a vowel is expressed in terms of a branching vs. non-branching nucleus, the default assumption is that diphthongs and vowels can, in principle, pattern phonologically in the same way as in (2), regardless of the phonetic details. This is the type of interpretation implied in the traditional descriptions of Old English (see also Lass 1994: 45–48), and its rejection as a theoretical possibility would have to derive from some – yet to be proposed – independently well-motivated principle. (2)
Vowel/diphthongal length contrast a. Short
b. Long
μ
μ
N
N
e
e
μ
μ
μ
N o
e
μ N
e
o
In sum, there is no doubt that the typological and theoretical issues relating to the Old English diphthongs are worthy of continued discussion, but in the meantime Hogg’s (1992b: 20) conclusion remains the most reasonable: “The evidence from both OE and ME suggests very strongly that the traditional position is in essence correct […]”.
3.3 Consonants With some exceptions, the general traits of the OE consonantal system are not particularly controversial. The basic inventory is provided in Table 17.2 (see Section 3.7 below for a set of phonological generalizations). In comparison with present-day standard varieties, there are differences in the inventory (for example, OE [c¸, x, ɣ]), suprasegmentals (almost all segments participate in a length contrast), and phonotactics (word edge clusters such as [kn-], [hl-], and [-mb], as in cne¯o ‘knee’, hlu¯d loud’, lamb ‘lamb’). One striking difference involves pairs of segments that are contrastive in modern varieties but only distributional variants in Old English (distributional pairs are boxed in Table 17.2). That is, the OE voiced fricatives [v, ð, z] occur only in a surrounding voiced environment – otherwise the voiceless counterparts occur – and [ŋ] occurs only preceding a velar. As is typically the case in writing systems, such phonetic detail was not usually indicated by the scribes, as in [f], [v] ‘wolf, wolves’. In the case of the velar voiceless fricative pair (both written as ), it is usually assumed that [c¸] occurred in a palatal environment ( ‘boy’), whereas [x] occurred elsewhere ( ‘brought’, ‘plough’). A reflection of this type of pattern can
17 Old English: Phonology
261
still be seen in modern standard German pairs such as ich [c¸] ‘I’ and acht [x] ‘eight’. Since the distribution of [h] (also written as ) was restricted (see Section 3.7, generalization l below), it is often treated as a distributional variant along with [c¸, x]. Further phonetic detail for individual segments beyond what is indicated in Table 17.2 can sometimes be reconstructed. For example, in the diphthongization process traditionally known as breaking (e > eo and a > ea triggered by specific post-vocalic consonants), the patterning of r with h [x] suggests a velarized variant of r in coda position, as in eorþe (< erþe) ‘earth’ and feohtan (< fehtan) ‘to fight’ (see Howell 1991). Particularly controversial areas of reconstruction involve the three segments in parentheses in Table 17.2; namely, [ ʃ ] (< [sk]), [tʃ ] (< [kj]), and [dʒ] (< [ɡj]). The difficulties lie not so much in the understanding of the general developments, but rather in attempting to determine a precise chronology of events. Table 17.2: Old English consonant system (late 10th century) Bilabial Labiodental Plosives
voiceless p voiced b
Fricatives
voiceless voiced
Nasals
t d f v
m
Inter- Alveolar Alveo- Palatal Velar Glottal dental palatal
θ ð
s z
(ʃ)
kj ɡj
k ɡ
c¸
x ɣ ŋ
n
Affricates
h
(tʃ ) (dʒ)
Liquids Lateral Central Approximants
l r j
w
In the case of [ ʃ ], there is no doubt of an assimilatory change involving coarticulatory effects by which early OE sk (usually written ) became [ ʃ ] by Middle English, a segment that is still maintained in Present-day English (ship < OE scip). The question that arises, though, is whether this change should be ascribed to the Old or early Middle English period. In fact, although [ ʃ ] is commonly reconstructed for Old English, the correctness of this assumption can be challenged in various ways. For example, in most OE poetry, as in Germanic generally, a cluster consisting of sp-, st-, or sk- can only self-alliterate. In late Old English, however, this strict system begins to break down, and there are cases of sp-, st-, and sk- alliterating with s-. Although the issues are complex, the alliteration of sk- with s- suggests that even in this later period we are still dealing with an sk- cluster, since alliteration of [ ʃ ] and [s] would be entirely unexpected (for detailed discussion, see Minkova 2003: 130–133). The evolution of the original voiceless velar plosive is a particularly problematic area. Again, from a bird’s eye diachronic perspective, the situation is relatively straightforward. We are dealing with phonological split arising from a very common type of sound change whose beginning and end points are clear; that is, a front vowel or j
262
III Old English environment results in a palatalized [kj], which ultimately evolves into an affricate, [tʃ ] (as in OE cild ‘child’ with original [k]). From a synchronic perspective, however, determining the chronology of the intermediate stages is a complex issue, and difficulties arise in reconstructing both the phonological status and the phonetic details of the evolving segments during the OE period. Although it is fair to say that the reconstruction of [tʃ ] already for early Old English has become a kind of standard theory (see, for example, Hogg 1992a: 95; Lass 2006: 54), in fact, robust evidence for the assumption of an affricate is strikingly absent. Indeed, the alliterative evidence once again suggests a more conservative progression of palatalization, since even in late OE verse, no distinction is made between k and the palatalized k for the purposes of alliteration, a practice that seems unlikely if palatalized k had already fully progressed to an affricate (for detailed discussion, see again Minkova 2003: 71–113, who argues further that the palatalized segment was not yet phonologized even in late Old English). The voiced velar obstruents also display a complex history, and the chronological and phonetic details are especially difficult due to the orthographic indeterminacy. The scribes did not consistently differentiate [ɡ], [ɣ], [ɡj], and [j], all of which could be written as (although the palatal quality could be indicated by adding an , as in [ɡj] ‘to singe’ and [j] ‘yoke’). In addition, the velar [ɡɡ] and palatal [ɡɡj] geminates could be written as or , as in ~ [ɡɡ] ‘dog’ and ~ [ɡɡj] ‘lay’. In a nutshell, early Old English displays the following phonological distribution of [ʝ], [ɣ], and [ɡ]: (a) the palatal fricative [ʝ] occurred in the front vowel environment (g(e)arn ‘yarn’, nægel ‘nail’), (b) the velar fricative [ɣ] occurred in a back vowel environment (ga¯st ‘spirit, ghost’, fugol ‘bird’), and (c) the plosive [ɡ] occurred only following a nasal and as a geminate (tunge ‘tongue’, dogga ‘dog’). By the classical OE period, the palatal fricative had merged with original [j], and [ɣ] had become [ɡ] word initially, so [ɣ] only occurred in the back vowel environment. Palatalized [ɡj] was found in the environment between a nasal and an original front vowel or j (seng(e)an < *sangjan ‘to singe’) or as the result of earlier West Germanic gemination (see OE lecgan vs. Go. lagjan ‘to lay’). This segment ultimately evolved into the affricate [dʒ] by Middle English, but for Old English an intermediate stage of palatalization can be assumed (represented here as [ɡj], in parallel with the case of [kj] discussed above). Even some modern dialects have forms such as brig ‘bridge’ and rig ‘ridge’ (OE brycg and hrycg), which suggest that in these dialects at least the affrication stage was never reached (Wright and Wright 1925: Section 319, note). In sum, the segments [ ʃ ], [tʃ ], and [dʒ] arguably do not belong to the OE inventory (hence their parenthetic status in Table 17.2), and in the case of the palatalized plosives it is plausible to assume an intermediate stage, which we represent here as [kj] and [ɡj].
3.4 Stress Although the ultimate details of OE stress are a complex matter, the location of primary word stress is relatively straightforward. In fact, for all words belonging to a major lexical category, except verbs, the primary generalization is simply: stress the initial syllable (regardless of whether it is a prefix or root syllable), as in a´nd-saca ‘adversary’, wæ´ ter ‘water’, ma´nigum ‘many-DAT.SG’, and also compounds such as bry¯´dguma ‘bridegroom’. At the same time, though, morphological factors come into play, and certain prefixes, such as ge- and be-, are never stressed, as in gewı´der ‘storm’ and bega´ng ‘practice’.
17 Old English: Phonology Verb stress, however, is somewhat more complicated. Prefixes are ignored for stress purposes – so the generalization here is: stress the initial root syllable of a verb, as in ofwu´ndrian ‘to be astonished’, on-sa´can ‘to dispute’ – unless the prefix is an adverbial or the verb derives from a noun carrying initial stress, in which case the stress falls on the prefix, as in ´ıncuman ‘to come in’, a´ndswarian ‘to answer’, cf. a´ndswaru ‘answer’. Although traditionally it is assumed that such verbal prefixes receive primary stress, Minkova (2003: 24–34) argues that stress on the prefix is subordinate to that of the root syllable. In fact, the topic of secondary stress in morphologically complex forms is a particularly difficult chapter of OE studies, especially since our primary source of evidence for stress patterns derives from the study of the complex interplay of stress and the metrical conventions of Old English poetry. For detailed discussion, see Lass 1994: 83–95 and Minkova 2006.
3.5 Unstressed syllables The segmental inventories presented in Section 3.2 and Section 3.3 above reflect the full set of contrasts found in the stressed environment. In many languages, unstressed syllables tolerate less complexity than stressed syllables. For example, unstressed syllable heads in Old English tend to be more restricted than stressed ones, so clusters such as kn- and hl- do not occur in the unstressed environment. In the case of vowels, unstressed syllables do not allow a vowel length contrast (although such a contrast can be reconstructed for pre-Old English; see Section 3.6 below). In addition, increasing vowel quality restrictions develop in unstressed syllables throughout the OE period. Although for classical Old English it is still possible to identify five vowels in unstressed syllables (that is, i, e, a, o, u; original æ had merged with e, and diphthongs are not possible), many mergers had in fact already taken place in specific environments. The general diachronic trend is for high vowels to lower to mid vowels, with local environmental factors facilitating or hindering the process. For example, -i becomes -e word finally, as in wine (< *wini) ‘friend’, but is preserved preceding palatal consonants and -ng, as in hefig ‘heavy’ ( = [j]) and cyning ‘king’. Similarly, -u generally becomes -o, although specific environments – for example, a following m – favor retention of u, as in heofon (< heofun) ‘heaven’ and fato (< fatu) ‘vats’, but sunum ‘son-DAT.PL’. Synchronically, given the relatively straightforward distribution of i ~ e, some argument can be made for treating them as distributional variants in unstressed syllables, but this is less likely in the case of u and o (see Hogg 1992a: 88). An important general issue here, however, is the possible effect of orthographic conservatism – particularly in the classical Old English manuscripts – since the written forms likely lag behind changes in pronunciation and do not reflect the full extent of phonological reduction (see Hogg 1992a: 121). Regardless, although once again it is difficult to determine with precision the Old English intermediate stages, the original vowel contrasts were neutralized in unstressed syllables by early Middle English, as reflected in the orthographic conflation of suffixal vowels to , as in name (OE nama) and tale (OE talu). Old English, like other early Germanic languages, allowed liquids and nasals to form the nucleus of an unstressed syllable, as in [hrı¯ðr̩ ] ‘head of cattle’ and [a¯dl̩] ‘disease’ (similarly, Go. akrs [akr̩ s] ‘acre’ and tagl [taɣl̩] ‘hair’). The diachronic source of such forms is found in vowel loss and, judging from OE poetry, the reduced forms were originally monosyllabic (that is, *akr < *akra- ‘acre’). However, they
263
264
III Old English regained a syllable probably first through the nuclearization of the liquid or nasal with later insertion of an anaptyctic vowel -i (later e) if the preceding vowel was front, and u (later o) if the preceding vowel was back. In fact, determining whether the nucleus of the unstressed syllable consisted of a sonorant or an anaptyctic vowel is difficult. Orthographic , for example, could be adopted by convention to represent nuclear l, just as writing might simply reflect a conservative use of orthography and not necessarily the absence of an anaptyctic vowel; thus [a¯tr̩ ~ a¯tor] ‘poison’, [næ¯dl̩] ~ [næ¯del] ‘needle’, [wæ¯pn̩ ~ wæ¯pen] ‘weapon’ (Campbell 1959: 151, Wright and Wright 1925: 100f). However, using the – not necessarily dependable – orthographic trends as a guide, it would appear that l, m, and n, which are often written in syllables without a vowel, were most likely to be nuclear. By contrast, r is seldom written alone, suggesting consistent anaptyxis. This would also reflect the pattern of earlier nuclearization, since the metrical evidence suggests that r was the first sonorant to nuclearize (Sievers 1893: Section 79,4a; Hogg 1992b: 237).
3.6 Quantity Old English, like the other early Germanic languages, is classified as a quantity language. Although all quantity languages – by definition – distinguish syllables in terms of weight, the phonological details of the weight contrast can vary from language to language. In fact, a common type of contrast is displayed in early Germanic: a codaless syllable with a short nucleus is light (monomoraic), all other syllables are heavy (bi- or polymoraic). This distinction is reflected in both sound change and poetic meter. For example, pre-Old English apocope of high vowels occurred after a heavy syllable but not after a light syllable, as in de¯or (< *de¯o.ru) ‘deer-PL’ but su˘nu (< *su˘.nu) ‘son’. In the case of Old English poetry, a property called resolution is displayed in which a sequence consisting of a light stressed syllable plus any immediately following syllable is treated as equivalent to a single heavy syllable, thus bisyllabic forms such as scipu ‘ships’ and werod ‘army’ can fill the same metrical position as heavy monosyllables such as wı¯f ‘woman’ and word ‘word’ (see, for example, Russom 1987: 12). In addition, quantity languages typically contrast both vowel and consonant length, but again there can be significant language-specific differences in the interplay of length, syllable structure, and stress. For example, a quantity language such as Finnish demonstrates complete independence of the three variables: the vowel length contrast is found in both stressed and unstressed syllables and in open and closed syllables, and geminate consonants can follow a long or short vowel, regardless of whether the vowel’s syllable is stressed or not (Becker 1998: 61–65). “Syllable-based quantity language” is the cover term for languages in which the weight contrast is found in both accented and unaccented syllables. Although Proto-Germanic – as well as the Indo-European classical languages such as Latin – belonged to this type, the attested early Germanic languages reflect a very strong tendency to begin restricting the full set of length contrasts and the weight contrast itself to stressed syllables only, leading to the designation “stress-based quantity language” (for full discussion, see Vennemann 1994, 1995). Late Old English, in its complete elimination of vowel and consonant length contrasts from unstressed syllables, had moved much further along this path than other Germanic languages such as Gothic and Old High German; for example, OE mihtı˘g, Go. mahteigs
17 Old English: Phonology ( = [ı¯]), OHG mahtı¯g ‘mighty’; OE sealfu˘de ~ sealfo˘de, Go. salbo¯da, OHG salbo¯ta ‘I anointed’ (see also Section 5.2 below). In sum, Old English can be described as a stress-based quantity language with morphologically-determined stress assignment (for further theoretical discussion, see also Dresher and Lahiri 1991 and Hayes 1995).
3.7 Phonological generalizations We present here a set of phonological generalizations that includes some of the main properties discussed above, as well as some other properties that distinguish Old English from present-day standard varieties. It is intended to be representative, not exhaustive. Below read “following” as “immediately following.” Standard orthographic forms are used (see Section 4). (For similar generalizations for Gothic, see Vennemann 1985.) a. Length contrasts occur only in a stressed environment: hw毴 las ‘whales’, 毴 las ‘eels’; -be´de ‘prayer-DAT’, be´dde ‘bed-DAT’. b. All segments except voiced fricatives, approximants, [ŋ], and [h] participate in the length contrast. c. Long consonants occur only in intersonorant environment following a stressed, short vowel: cynnes ‘kin-GEN’, bettra ‘better’. d. Long consonants are geminates; that is, they close and give weight to the preceding syllable; pyffan [pyf.fan] ‘puff’, dogga [doɡ.ɡa] ‘dog’. e. [v, ð, z] occur only in a surrounding voiced environment: seofon ‘seven’, wulfas ‘wolves’, lifde ‘s/he lived’. f. Non-geminate [f, θ, s] do not occur in a surrounding voiced environment (see e). g. Non-geminate [b] occurs only word-initially or following a nasal: blo¯d ‘blood’, climban ‘to climb’. h. Non-geminate [ɡ] occurs only following a nasal: singan ‘to sing’. In late Old English, it also occurs word initially; gu¯þ ‘combat, war’. i. [ɣ] does not occur following a nasal or a front vowel (compare fugol ‘bird’, swelgan ‘to swallow’, both with [ɣ]). In late Old English, it also does not occur word initially (see h). j. [j] does not occur following a back vowel (compare dæg˙ [j] ‘day’). k. Non-geminate [ɡj] occurs only following [n] and word finally: seng(e)an ‘singe’, ecg ‘edge’. l. [h] occurs only in word initial position or following certain prefixes such as be- and ge-: hof ‘enclosure, court’, behindan ‘behind’, geheald ‘keeping custody’. m. [c¸], [c¸c¸] occur only following a front vowel or diphthong, and non-geminate [c¸] occurs only in syllable coda: hliehhan ‘to laugh’, cnihtas ‘boys’, riht ‘right’. n. [x] does not occur word initially, and [x], [xx] do not occur following a front vowel or diphthong (see (m); compare pohha ‘pocket’, dohtor ‘daughter’, to¯h ‘tough’, sulh ‘plow’, all with [x]). o. [ŋ] occurs only preceding a velar: drincan ‘to drink’, singan ‘to sing’. p. The sequences below can form syllable heads (onsets) under stress: [k] or [h] plus one of [n, l, r, w]: cna¯wan ‘to know’, clæ¯ne ‘pure, clean’, cre¯da ‘belief, creed’, cwe¯n ‘woman, queen’; hnutu ‘nut’, hlu¯d ‘loud’, hring ‘ring, fetter’, hwæt ‘what’. [ɡ] plus one of [n, l, r]: gnagan ‘to gnaw’, glı¯dan ‘to glide’, gre¯tan ‘to greet’. [f] plus one of [n, l, r]: fne¯san ‘to sneeze’, flo¯d ‘flood’, friþa ‘protector’.
265
266
III Old English [w] plus one of [l, r]: wlitig ‘radiant, beautiful’, wræc ‘misery’. q. [mb], [nɡ] can form syllable codas: lamb ‘lamb’, lang ‘long’. r. Only vowels and diphthongs can form the nucleus of a stressed syllable. s. Only (short) vowels (specifically [i, e, a, o, u]), m, n, l, and probably r can form the nucleus of an unstressed syllable: ma¯þm̩ ~ ma¯þum ‘gift’, be¯acn̩ ~ be¯acen ‘sign, beacon’, segl̩ ~ segel ‘sail’, a¯tr̩ ~ a¯tor ‘poison’.
4 Phonological–orthographic correspondences As mentioned above, for the most part OE orthography is phonologically based and, in fact, most graphs can be roughly interpreted in terms of their equivalent IPA values; for example, , represent [æ], [y], respectively, as in fæder ‘father’ and hyll ‘hill’. At the same time, a consistent indication of vowel and diphthongal length is notably absent in the manuscripts, so , for example, represents both the short and long vowel; [ky̆ niŋɡ] ‘king’, [bry¯d] ‘bride’. In addition, orthographic geminates are often maintained word finally, although phonological geminates are not found in this position, as in bedd [bed], bedde [bedde] ‘bed-NOM/DAT’. Some primary conventions that are not transparent or self-resolving are listed below. Although we abstract away from the significant variation evident in the manuscripts, the correspondences given in Table 17.3 provide a reasonable reflection of classic OE practices. (We include [ø], although in fact it underwent early derounding in West Saxon; see Section 3.1 above.) Table 17.3: Old English phonological–orthographical correspondences Phonology
Orthography
Example
Phonological Form
Gloss
(a)
[f] ~ [v]
(b)
[ff] [θ] ~ [ð]
~
(c)
[θθ] [s] ~ [z]
~
(d)
[ss] [k]
, also
(e)
[kj] (> [tʃ ])
, also
(f)
[ɡ] ~ [ɣ]
wulf wulfe pyffan þencan ~ ðencan broþor ~ broðor moððe ~ moþþe sæ nosu is cyssan cræft weorc kyning cinn bec þeccean ʒast brinʒan laʒu
wulf wulve pyffan θeŋkjan bro¯ðor moθθe sæ¯ nozu ı¯s kyssan kræft weork kyniŋɡ kjinn be¯kj θekkjan ɡa¯st briŋɡan laɣu
‘wolf’ ‘wolf-DAT’ ‘puff’ ‘to think’ ‘brother’ ‘moth’ ‘sea’ ‘nose’ ‘ice’ ‘to kiss’ ‘skill’ ‘work’ ‘king’ ‘chin’ ‘books’ ‘to cover ‘spirit’ ‘to bring’ ‘law’ (Continued )
17 Old English: Phonology
267
Table 17.3: Continued Phonology
Orthography
(g)
[j]
(h) (i) ( j) (k)
[ɡj] (> [dʒ]) [ɡɡj] (> [(d)dʒ]) [ɡɡ] [h], [x] ~ [c¸]
(l)
[n] ~ [ŋ]
(m)
[w]
(n)
[æ¯̆ ]
(o)
[y¯̆ ]
(p)
[ø̄̆ ]
(q)
[æ¯̆ a̯ ]
(r)
[e¯̆ o̯ ]
(s)
[ks]
ʒift ʒeong dæʒ , also brycʒ ~ lecʒan ~ leʒʒan ~ froʒʒa ~ frocʒa
hamor behindan dohtor crohha flyht
hnutu spinnan tunʒol
ƿolcen saƿol
fæstan sæd
þyncan yþ
oele cwoen
eall deaþ
ʒeolo deop
axian , also
Example
Phonological Form
Gloss
jift joŋɡ dæj briɡj leɡɡjan froɡɡa hamor behindan doxtor kroxxa flyc¸t hnutu spinnan tuŋɡol wolkn̩ sā wol fæ̆ stan sæ¯d θy̆ ŋkjan y¯θ ø̆ le kwø̄ n æ̆ a̯ l dæ¯a̯ θ je˘o̯ lo de¯o̯ p a¯ksian
‘marriage gift’ ‘young’ ‘day’ ‘bridge’ ‘to lay’ ‘frog’ ‘hammer’ ‘behind’ ‘daughter’ ‘crock pot’ ‘flight’ ‘nut’ ‘to spin’ ‘star’ ‘cloud’ ‘soul’ ‘to fast’ ‘seed’ ‘to seem’ ‘wave’ ‘oil’ ‘queen’ ‘all’ ‘death’ ‘yellow’ ‘deep’ ‘to ask’
The scribes typically did not make any phonological distinction in their use of and (Hogg 1992b: 33–34). The graph was occasionally used instead of , especially preceding , as in kyning ‘king’. If the sound change *sk > [ ʃ ] is posited for Old English (see Section 4.3 above), then [ ʃ ] is represented by . By the time of Middle English, [ ʃ ] was usually written or . For the diphthongs note that we give here a literal transcription ([eo̯ and [æa̯ ]), although various off-gliding values can be assumed (see Section 3.2). Old English texts are subjected to modern editorial conventions. For example, although the symbol was used by OE scribes, most modern works transliterate using . Note, however, that by early Middle English and were used contrastively for [j] and [ɡ], respectively. Similarly, the symbols (for [kj], later [tʃ]), (for [j]), and / (for [(ɡ)ɡj], later [dʒ]), are not found in the manuscripts; they are used by modern editors to distinguish the palatal and velar counterparts; see (d)–(j) above. Finally, although both and were used in early texts, in fact [w] was typically represented by runic wynn, ƿ. In modern works, the rune symbol is invariably transliterated as .
268
III Old English
5 Diachrony 5.1 The “age of harmony”: Umlaut No outline of Old English phonology would be complete without mention of what Lass (1994: 59) labels “the age of harmony”, which subsumes the various types of umlaut (vowel harmony changes) that took place from the Proto-Germanic to early Old English periods. Germanic umlaut is a type of partial regressive assimilation in which the target vowel – typically the stressed vowel – takes on qualities of a following trigger vowel. Umlaut is pervasive in all the early Germanic languages with the exception of Gothic, and its effects include all the logical possibilities of lowering, backing, fronting, and raising. Lowering of u to o, for example, is evident in West Germanic a-umlaut, as in OE gold (< *gulda) ‘gold’. Backing of *æ to a in the environment of a following back vowel is found in later pre-Old English in a change that is often called Restoration, since preOE æ, which arose through an earlier general fronting of original a, was “restored” to a under back umlaut conditions, as reflected, for example, in paradigmatic allomorphy of the type dæg, dagas ‘day, days’ and fæt, fatu ‘vat, vats’. Back umlaut can also produce diphthongs (assuming the existence of short diphthongs; see Section 3.2 above), although this change had only limited effect in West Saxon; eofor (< efor) ‘boar’, heorut (< herut) ‘hart’. In Old English, the most generalized and systematic subtype is i-umlaut, which involves primarily the fronting of vowels and diphthongs under the influence of a following i or j, although raising in the case of the short low vowel can also occur (see Table 17.4). (Note that e had already been raised to i in earlier Germanic. Also, umlauted a˘ preceding a nasal is found as æ̆ in early texts, and then usually e˘ in later ones.) Table 17.4: Old English i-umlaut (vowels)
æ̆ > e˘ ă > æ̆ > e˘ a¯ > æ¯ o˘ > ø̆ (> e˘) o¯ > ø̄ (> e¯) u˘ > y̆ u¯ > y¯
i
y˘¯
e˘
ø˘¯
u˘¯ o˘¯
˘ æ
a˘
æ¯
a¯
bedd senda dæ¯lan dehter se¯can cynn dy¯stig
(Go. badi) (Go. sandjan) (*da¯ljan) (*dohtri) (Go. so¯kjan) (Go. kuni) (compare du¯st ‘dust’)
‘bed’ ‘to send’ ‘to divide’ ‘daughter-DAT’ ‘to seek’ ‘race, generation’ ‘dusty’
In general, i-umlaut in Old English does not display the complexity it does in the other Germanic languages. In Old High German, for example, the intervening consonantal environment plays a significant role in facilitating or hindering umlaut, and in Old Norse there is a complex interaction between umlaut and other sound changes, especially i-syncope (Howell and Salmons 1997; Iverson and Salmons 2004). By contrast, the West Saxon situation reflects a relatively straightforward, highly generalized application in which long and
17 Old English: Phonology short u, o, and a regularly undergo umlaut. At the same time, though, there are two primary restrictions. First, raising umlaut of short æ (> e) can be blocked by (non-geminate) clusters; for example, umlaut is found in hebban ‘to raise’ but not in fæstan ‘to make firm’ (both with an original *-jan suffix). Second, long æ¯ resists umlaut altogether, as in læ¯ce ‘physician’, where æ¯ remains in spite of the original *-ja- (> e) suffix. In fact, these restrictions constitute a pattern in conformance with other Germanic languages and general principles governing umlaut. Howell and Salmons (1997: 89) show that the propensity to undergo umlaut increases in accordance with the degree of qualitative difference between trigger and target vowels, which in the case of Old English relates primarily to the back–front dimension. Thus, although fronting of the back vowels (u, o, a) is regular, short æ, which is already front, can resist raising umlaut in the cluster environment. Similarly, the even greater resistance to umlaut displayed by the long vowel æ¯ is typical, as umlaut preferentially affects short vowels. For example, in Old High German, short a is most susceptible to umlaut (known as primary umlaut), and in Dutch only short vowels undergo umlaut. These assimilatory changes had a significant impact on OE phonology and morphology. At first, the umlaut vowels [ø] and [y] were only distributional variants of o and u occurring under specific conditions, but already in pre-Old English they were phonologized, yielding two new segments. Although ø underwent early derounding in West Saxon, y was relatively stable. Along with this phonologization came a dramatic increase in allomorphy, which is evident throughout the lexicon; for example hnutu, hnyte ‘nut-NOM, nut-DAT’, ic do¯, he de¯þ ‘I do, he does, bra¯d, bræ¯dra ‘broad, broader’. Although most of the allomorphy was leveled out in later stages, some traces remain even into Present-day English. Modern pairs displaying lexical split such as older, elder and brothers, brethren have developed in accordance with Kuryłowicz’s (1947) fourth “law” governing leveling; that is, the original form (umlauted elder, brethren) took on a specialized meaning, while the new form carries the primary meaning. Finally, i-umlaut was morphologized as a plural marker, although only a handful of these umlaut plurals remain in present day English, as in man, men; foot, feet; and mouse, mice (compare OE a¯c, æ¯c ‘oak, oaks’; bo¯c, be¯c ‘book, books’, and cu¯, cy¯ ‘cow, cows’).
5.2 Quantity changes Both the pre-Old English and late Old English/early Middle English stages were robust periods of quantity change. The early Germanic and pre-Old English changes primarily affected unstressed syllables, while stressed syllables were targeted in the later changes. In comparison with these early and late stages, the OE period itself was relatively stable, especially with regard to the stressed syllable. The pre-Old English period involved a severe reduction of unstressed syllables, including various kinds of vowel and consonant deletions and loss of vowel length, as in bend (Ø < -i; compare Go. bandi ‘band, ribbon’); hand (Ø < -uz; Go. handus ‘hand’), and sealfude (u˘ < o¯; Go. salbo¯da ‘he anoints’). Traditionally, this reduction is commonly linked to the shift from pitch to stress accent and the fixing of stress on the root syllable that occurred in Proto-Germanic. In addition, although difficult to quantify, it is also sometimes claimed that the intensity of the stress accent gradually increased throughout the history of English with concomitant weakening of unstressed syllables. (See Luick 1914–21: 267–362; Lass 1994: 95–102; Lutz 1991: 281–282.) In fact, the amount of diachronic reduction that unstressed syllables undergo can vary significantly from one language to the next. For example,
269
270
III Old English Finnish (with its initial stress accent) is extremely conservative in maintaining contrasts in unstressed syllables, a fact strikingly apparent in its treatment of very early borrowings from Germanic; for example, Finnish kuningas, PGrmc. *kuningaz ‘king-NOM’. (For further discussion, see van Coetsem et al. 1981; Salmons 1992: 166–168; Boutkan 1995.) The late Old English and Middle English changes are particularly significant, since they involve the complete breakdown of the original quantity system, an event that occurred in almost all Germanic dialects at one point or another during the medieval period. Three of the main quantity changes assumed are Closed Syllable Shortening (CSS), which eliminated the vowel length contrast in closed disyllables (ke¯pte > ke˘pte ‘kept’); Degemination, which eliminated the contrast between short and long consonants (OE æppel [pp], ME apel [p], where usually an ambisyllabic consonant is assumed for the latter); and Open Syllable Lengthening (OSL), which eliminated the vowel length contrast in open syllables (OE na˘ma, ME na¯me ‘name’). Although these changes are properly ascribed to the late Old English (CSS) and Middle English periods (Degemination, OSL), Luick’s (1898, 1964a [1914–21]) view that they represent a continuation of a process that began in preOld English, and even in West Germanic, has been extremely influential (see, for example, Ritt 1994). According to Luick (1898), there was a rhythmic tendency operative throughout early English in which syllable weight was gradually being standardized according to a set of prosodic weight templates. For example, in the case of disyllables, the ideal stressed syllable was assumed to be bimoraic. Thus, CSS (through shortening and loss of a mora) and OSL (through lengthening and addition of a mora) yielded this ideal type (ke˘p.te, na¯.me), while syllables already bimoraic remained unchanged. In fact, though, for the Old English period, evidence for a standardization of quantity is not particularly robust, and in the late Old English period, at least one important change moved the stressed syllable away from the bimoraic ideal. In the change known as Homorganic Cluster Lengthening (see Section 3.1), in which vowels lengthened before clusters such as -ld, -mb, and -nd, an already bimoraic syllable became overlong, as in ME clı¯m.ben (< OE clı˘m.ban) ‘to climb’. Further, even in the case of Middle English, Degemination does not conform to the bimoraic preference, since the stressed syllable is already bimoraic at the pre-Degemination stage, as in OE æp.pel. For detailed discussion and an alternative interpretation, see Murray 2000 (and references there) and Mailhammer 2007.
6 References Becker, Thomas. 1998. Das Vokalsystem der deutschen Standardsprache. Frankfurt: Peter Lang. Brunner, Karl. 1965. Altenglische Grammatik. Nach der angelsa¨chsischen Grammatik von Eduard Sievers. 3rd edn. Tu¨bingen: Max Niemeyer. Boutkan, Dirk. 1995. The Germanic “Auslautgesetze”. Amsterdam: Rodopi. Bu¨lbring, Karl D. 1902. Altenglisches Elementarbuch. Vol. I: Lautlehre. Heidelberg: Winter. Campbell, Alistair. 1959. Old English Grammar. Oxford: Oxford University Press. Daunt, Marjorie. 1939. Old English sound changes reconsidered in relation to scribal tradition and practice. Transactions of the Philological Society 1939 38: 108–137. Dresher, B. Elan and Aditi Lahiri. 1991. The Germanic foot: Metrical coherence in Germanic. Linguistic Inquiry 22(2): 251–286. Fox, Anthony. 1995. Linguistic Reconstruction: An Introduction to Theory and Method. Oxford: Oxford University Press. Hayes, Bruce. 1995. Metrical Stress Theory: Principles and Case Studies. Chicago: University of Chicago Press.
17 Old English: Phonology Hockett, Charles F. 1959. The stressed syllabics of Old English. Language 35: 575–597. Hogg, Richard M. 1992a. Phonology and morphology. In: Richard M. Hogg (ed.), The Cambridge History of the English Language. Vol. I. The Beginning to 1066, 67–167. Cambridge: Cambridge University Press. Hogg, Richard M. 1992b. A Grammar of Old English. Oxford: Blackwell. Hogg, Richard M. 2006. Old English Dialectology. In: van Kemenade and Los (eds.), 395–416. Howell, Robert B. 1991. Old English Breaking and its Germanic Analogues. Tu¨bingen: Max Niemeyer. Howell, Robert B. and Joseph C Salmons. 1997. Umlautless residues in Germanic. American Journal of Germanic Linguistics 9: 83–111. Iverson, Gregory K. and Joseph C Salmons. 2004. The conundrum of Old Norse umlaut: Sound change versus crisis analogy. Journal of Germanic Linguistics 16: 77–110. Kemenade, Ans van and Bettelou Los (eds.). 2006. The Handbook of the History of English. Oxford: Blackwell. Kuryłowicz, Jerzy. 1947. La nature des proce`s dits analogiques. Acta Linguistica 5: 15–37. (Reprinted in Eric P. Hamp, Fred W. Householder, and Robert Austerlitz (eds.), Readings in Linguistics 2, 158–174. Chicago: Chicago University Press, 1966.) Lass, Roger. 1992. Phonology and morphology. In: Norman Blake (ed.), The Cambridge History of the English Language. Vol. II. 1066–1476, 23–155. Cambridge: Cambridge University Press. Lass, Roger. 1994. Old English: A Historical Linguistic Companion. Cambridge: Cambridge University Press. Lass, Roger. 2006. Phonology and morphology. In: Richard Hogg and David Denison, A History of the English Language, 43–108. Cambridge: Cambridge University Press. Lass, Roger and John M. Anderson. 1975. Old English Phonology. Cambridge: Cambridge University Press. Luick, Karl. 1898. Beitra¨ge zur englischen Grammatik III: Die Quantita¨tsvera¨nderungen im Laufe der englischen Sprachentwicklung. Anglia 20: 335–362. Luick, Karl. 1964a [1914–1921]. Historische Grammatik der englischen Sprache, Vol. 1, Part 1. Stuttgart: Bernhard Tauchnitz. Luick, Karl. 1964b [1929–1940]. Historische Grammatik der englischen Sprache, Vol. 1, Part 2. Stuttgart: Bernhard Tauchnitz. Lutz, Angelika. 1991. Phonotaktisch gesteuerte Konsonantenvera¨nderungen in der Geschichte des Englischen. Tu¨bingen: Max Niemeyer. Maddieson, Ian. 1984. Patterns of Sounds. Cambridge: Cambridge University Press. Mailhammer, Robert. 2007. On syllable cut in the Orrmulum. In: Christopher M. Cain and Geoffrey Russom (eds.), Studies in the History of the English Language III – Managing Chaos: Strategies for Identifying Change in English, 37–61. Berlin/New York: Mouton de Gruyter. Minkova, Donka. 2003. Alliteration and Sound Change in Early English. Cambridge: Cambridge University Press. Minkova, Donka. 2006. Old and Middle English prosody. In: van Kemenade and Los (eds.), 95–124. Minkova, Donka and Robert P. Stockwell. 1992. Homorganic clusters as moric busters in the history of English: The case of -ld, -nd, -mb. In: Matti Rissanen, Ossi Ihalainen, Terttu Nevalainen and Irma Taavitsainen (eds.), History of Englishes: New Methods and Interpretations in Historical Linguistics, 191–207. Berlin/New York: Mouton de Gruyter. Murray, Robert W. 2000. Syllable cut prosody in Early Middle English. Language 76: 617–681. Noss, Richard B. 1964. Thai Reference Grammar. Washington, DC: Foreign Service Institute. Pyles, Thomas and John Algeo. 1993. The Origins and Development of the English Language. 4th edn. Orlando, FL: Harcourt Brace Jovanovich. Ritt, Nikolaus. 1994. Quantity Adjustment: Vowel Lengthening and Shortening in Early Middle English. Cambridge: Cambridge University Press. Rosner, B. S. and J. B. Pickering. 1994. Vowel Perception and Production. Oxford: Oxford University Press.
271
272
III Old English Russom, Geoffrey. 1987. Old English Meter and Linguistic Theory. Cambridge: Cambridge University Press. Salmons, Joseph. 1992. Accentual Change and Language Contact: Comparative Survey and Case Study of Early Northern Europe. Stanford: Stanford University Press. Sievers, Eduard. 1893. Altgermanische Metrik. Halle: Max Niemeyer. Sievers, Eduard. 1898 [1882]. Angelsa¨chsische Grammatik. Halle: Max Niemeyer. Sievers, Eduard. 1901. Grundzu¨ge der Phonetik zur Einfu¨hrung in das Studium der Lautlehre der indogermanischen Sprachen. 5th edn. Leipzig: Breitkopf & Ha¨rtel. Stockwell, Robert P. 1952. Chaucerian Graphemics and Phonemics: A Study in Historical Methodology. Unpublished doctoral dissertation. University of Virginia, Charlottesville. Stockwell, Robert P. and C. Westbrook Barritt. 1951. Some Old English Graphemic-Phonemic Correspondences - ae, ea and a. (Studies in Linguistics, Occasional Papers, 4.) Norman, OK: Battenburg Press. Stockwell, Robert P. and C. Westbrook Barritt. 1955. The Old English short digraphs: Some considerations. Language 31: 373–389. Stockwell, Robert P. and Donka Minkova. 1990. The Early Modern English vowels, more o’ Lass. Diachronica 7: 199–214. Sweet, Henry. 1888. History of English Sounds. 2nd edn. Oxford: Clarendon Press. van Coetsem, Ronald, Hendricks and Susan McCormick. 1981. Accent typology and sound change. Lingua 53: 295–315. Vennemann, Theo. 1985. Phonologically conditioned morphological change: Exceptions to Sievers’ Law in Gothic. In: Edmund Gussman (ed.), Phonomorphology: Studies in the Interaction of Phonology and Morphology, 193–219. Lublin: Katholische Universita¨t Lublin. Vennemann, Theo. 1994. Universelle Nuklearphonologie mit epipha¨nomenaler Silbenstruktur. In: Karl Heinz Ramers, Heinz Vater, and Henning Wode (eds.), Universale phonologische Strukturen und Prozesse, 7–54. Tu¨bingen: Max Niemeyer. Vennemann, Theo. 1995. Der Zusammenbruch der Quantita¨t im Spa¨tmittelalter und sein Einfluß auf die Metrik. Quantita¨tsproblematik und Metrik. Amsterdamer Beitra¨ge zur a¨lteren Germanistik (special volume ed. by Hans Fix) 42: 185–223. White, David L. 2004. Why we should not believe in short diphthongs. In: Anne Curzan and Kimberly Emmons (eds.), Studies in the History of the English Language II. Unfolding Conversations, 57–84. Berlin/New York: Mouton de Gruyter. Wright, Joseph and Elizabeth Mary Wright. 1925. Old English Grammar. Oxford: Oxford University Press.
Robert Murray, Calgary (Canada)
18 Old English: Morphology 1 Inflectional morphology of the noun phrase 2 Inflectional morphology of verbs 3 References
Abstract Old English is in many respects a typical Indo-European language. This is particularly true of its morphological categories and its complex inflectional systems. It is mainly due to this complexity that this chapter cannot treat all aspects of OE Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 272–293
18 Old English: Morphology morphology in full detail. It therefore focuses on the most important inflectional systems of Old English. Morphological word-formation patterns are necessarily treated only marginally (Kastovsky, Chapter 9). Moreover, there is a considerable degree of dialectal variation in Old English which is also manifested in the morphological paradigms. This variation cannot be covered here comprehensively. This article therefore has a strong bias towards the later stages of the West Saxon variety – the dialect and period from which the greatest share of our extant sources is transmitted. For more comprehensive accounts, including the details of the diachronic and diatopic variation, I refer the reader to the relevant sections in Hogg and Fulk (2011) as well as to the older, but still valuable works by Campbell (1959) and Brunner (1965).
1 Inflectional morphology of the noun phrase 1.1 The inflectional categories of the noun phrase Case/number is marked by inflection (as opposed to agglutination), that is, there is no distinctive marker encoding a value of only one of the two categories. Case/number is generally marked on any element of the noun phrase, i.e. on all modifiers and on the head noun. The noun paradigms are the least distinctive of all, so that in many instances only the case marking on the adjective and/or determiner can unambiguously indicate the case/number value of the entire noun phrase. Old English (OE) has a typically Indo-European (IE) case system. Therefore any general descriptions of the functions of IE cases can well be applied to Old English. Of the eight IE cases, five survived: nominative, accusative, dative, genitive, and instrumental. The former three are used predominantly to mark grammatical relations. The genitive case marks a noun phrase as a modifier of a superordinate noun phrase. The instrumental is formally distinguished from the dative only in some adjectival and pronominal paradigms. If it is encoded, it marks complements that take the semantic role of an Instrument. For good descriptions of the functions of the five OE cases cf. Blake (2001: Section 2.3) or, specifically for Old English, Mitchell (1985: I, Sections 1240–1427). There are two number values for English nouns: singular and plural. A third number value, the dual, can be encoded in some pronominal paradigms. There are three gender values, all inherited from Proto-Indo-European, masculine, feminine, and neuter. The gender value of a noun determines agreement patterns on pronouns and on dependent adjectives and determiners so that, while inherent in the lexical entry of a noun, gender is morphologically encoded on any noun modifier. As in Indo-European languages in general, there is no distribution along the lines of natural gender. Particularly expressions denoting female persons can have masculine (wı¯fman ‘woman’) or neuter (wı¯f ‘woman’) gender. Expressions for inanimate referents do not necessarily have neuter gender. Only some expressions that may refer to either a female or a male person, such as e.g. names of occupations, can have both a masculine and a feminine form, which is then used to semantically distinguish between a female and a male member of that class (e.g. munuc ‘monk’ vs. mynecen ‘nun’).
273
274
III Old English
1.2 The inflectional paradigms of nouns In every noun class there is some degree of syncretism of forms. As in all IE languages, the following values are not distinguished formally in any noun class of Old English: nominative and accusative of neuter nouns for both number values, and the genitive plural and the dative plural of all three genders. The traditional labels of the major noun classes refer to the thematic morpheme reconstructed for Proto-Indo-European, which was inserted between the lexical root and the inflectional suffixes. Although the thematic elements “are not synchronically transparent and reflect the product of historical reconstructions” (Blake 2001: 4; cf. also Lass 1994: 123), they are still in common use in the description of Indo-European daughter languages. There is a major division into vocalic classes (classes 1 to 4) and consonantal classes (classes 5 and 6) according to whether the thematic element was a vowel or a consonant. Outside the vocalic/consonantal distinctions, there are other classes that originally did not have a thematic element and are therefore called “athematic” nouns. None of these terms refer to properties that are synchronically transparent in Old English. In the following, I simply specify the noun classes by numbers from 1 to 8; but I add the traditional labels in the section headings.
1.2.1 Class 1 – Germanic a-stems, Indo-European o-stems The nouns of class 1 are all either masculine or neuter. There are three main paradigms of this class, according to whether the noun is masculine (class 1a), neuter with a short root syllable (class 1b), or neuter with long root syllables, i.e., either with a long root vowel or with a V(ː)C-cluster in the nucleus of the root (class 1c); cf. Tables 18.1–18.3. Phonological cross-influences created a number of variant paradigms, an overview of which will be presented in the following. Table 18.1: Class 1a (sta¯n ‘stone’)
NOM ACC DAT GEN
SG
PL
sta¯n-∅ sta¯n-∅ sta¯n-e sta¯n-es
sta¯n-as sta¯n-as sta¯n-um sta¯n-a
Table 18.2: Class 1b (scip ‘ship’)
NOM ACC DAT GEN
SG
PL
scip-∅ scip-∅ scip-e scip-es
scip-u scip-u scip-um scip-a
18 Old English: Morphology
275
Table 18.3: Class 1c (word ‘word’)
NOM ACC DAT GEN
SG
PL
word-∅ word-∅ word-e word-es
word-∅ word-∅ word-um word-a
Apophonic variation: Because /ɛ/ was lowered to /a/ before palatal vowels in pre-Old English, nouns with the root vowel /ɛ/ show an apophonic alternation between the singular (uninflected or with suffixes containing a back vowel) and the plural forms (with suffixes all containing a front vowel). This applies to 1a and 1b nouns, so that there is a root variation as in dæg- ‘day-SG’, fæt- ‘vessel-SG’ vs. dag- ‘day-PL’, fat- ‘vessel-PL’. Root-final /-χ/: If followed by a vowel, root-final /-χ/ was lost. It was retained only in the uninflected forms, in which it was represented as 〈-h〉. In the relevant nouns we find, mearh-∅ ‘horse-NOM/ACC.SG’, but mear- for all other values. Syncope and epenthesis: Two types of paradigm-internal variation may look synchronically like one and the same phenomenon, but result from two opposing processes: there is both a syncope of an unstressed vowel in a bisyllabic root, e.g. engel-∅ ‘angel-NOM/ACC.SG’ vs. engl-es ‘angel-GEN.SG’, and a vowel epenthesis in a cluster consisting of a stop and a liquid, e.g. fugol-∅ ‘bird-NOM/ACC.SG’ formed from fugl- ‘bird’. The Germanic (Grmc.) ja-stems: This variant is considered a distinct subclass in Proto-Germanic. Originally, the thematic element preceding the inflectional suffixes was */-ja-/. In many forms originally belonging to this subclass, there are no differences from the default paradigms of classes 1a–1c in Tables 18.1–18.3. In some cases, however, the glide may still be represented as a root-final /-j-/ (usually spelled 〈-g-〉), or as /-ə/ following the root. Such traces can be found, for example, in her-e-∅ ‘army-THM-NOM/ACC.SG’ vs. her-g- (/ˈher-j-/) for all other values. (THM=thematic element.) The Grmc. wa-stems: Similar to the previous group there are some nouns descending from the PGrmc. sub-class with */-wa-/ as thematic element. It has been retained in some OE nouns as a root-final /-u/ in the uninflected forms and as a glide between root and suffix in the inflected forms. This is the case, for instance, in bear-u-∅ ‘grove-THM-NOM/ACC.SG’ vs. bear-w- for all other values.
1.2.2 Class 2 – Germanic o¯-stems, Indo-European a¯-stems The nouns of this class are all feminine. The nominative/accusative plural forms differ in West-Saxon from the other regional varieties of Old English. As in the neuter nouns of class 1, there is a distinction between nouns with short syllables (class 2a) and with long syllables (class 2b). In class 2a, later West Saxon (WSax.) texts show the genitive plural ending -ena for -a. Cf. Tables 18.4 and 18.5. Phonological crossinfluences are less numerous than those of class 1. The most important ones are briefly presented below:
276
III Old English Table 18.4: Class 2a (talu ‘tale’)
NOM ACC DAT GEN
SG
PL
tal-u tal-e tal-e tal-e
tal-a (WSax.), tal-e (non-WSax.) tal-a (WSax.), tal-e (non-WSax.) tal-um tal-a, tal-ena (late WSax.)
Table 18.5: Class 2b (wund ‘wound’)
NOM ACC DAT GEN
SG
PL
wund wund-e wund-e wund-e
wund-a (WSax.), wund-e (non-WSax.) wund-a (WSax.), wund-e (non-WSax.) wund-um wund-a
The Grmc. jo¯-stems: As in class 1, there is a subclass postulated for those nouns of class 2, in which the thematic element was */-joː-/. The glide is usually no longer present, but it caused the nominative singular suffix -u to be reduced to /-ə/ and to be dropped subsequently. Otherwise the paradigm does not differ from the default ones shown above in Tables 18.4 and 18.5. The Grmc. wo¯-stems: The thematic morpheme */-woː-/ is usually retained in inflected forms as /-u/ after short root syllables and before inflectional suffixes, for instance in sin-u-∅ ‘sinew-THM-NOM.SG’. In forms with an inflectional suffix it remained a glide, e.g. sin-w- ‘sinew-THM-’. After long root syllables the thematic element was dropped.
1.2.3 Class 3 – Germanic i-stems The forms of class 3 merged to a large extent with those of class 1a, class 2, and classes 1b or 1c, depending, respectively, on whether they are masculine, feminine, or neuter. The original thematic vowel of class 3, */-i-/, is no longer retained in most OE forms. Only in the nominative and accusative singular forms of masculine and neuter nouns with short syllables does final /-ə/ attest to the former i-suffix, e.g. win-e ‘friend-NOM/ACC.SG(MASC)’, sper-e ‘spear-NOM/ACC.SG(NEUT)’. However, the suffix left a trace in all members of this class as it caused i-umlaut in the root vowel. So, in spite of the fact that a number of inflectional suffixes were transferred from other noun classes, the umlauted root vowel constitutes the major formal difference between class 3 and classes 1 and 2. Yet, since there is no alternation of the root vowels within the paradigms of any subgroup of class 3, this is basically a diachronic feature and has no significance synchronically. Where there are original i-stem suffixes, they usually occur in doublets with corresponding forms of classes 1 and 2. For instance, in masculine nouns with short syllables inherited suffixes still compete with those of class 1. In the nominative and accusative plural we find both a reflex of the original -i (reduced to /-ə/) and the class 1 suffix -as. In the genitive plural, the original suffix /-ija/ (usually spelled 〈-iga〉) is still attested in poetry. This results in forms like win-e next to win-as ‘friend-NOM/ACC.PL’, win-iga (poetic) next to win-a ‘friend-GEN.PL’.
18 Old English: Morphology
277
In the plural forms of masculine nouns with long syllable the pattern is generally that of class 1a. Only exceptionally do we find suffixes of the original class 3 in some tribal names as e.g. Engl-e ‘Angle-NOM/ACC.PL’. In the paradigm of feminine nouns of class 3 with long root syllables, original forms in /-ə/ (< /-i/) and those of class 2a compete only in the accusative singular, so that both dæ¯d-∅ and dæ¯d-e ‘deed-ACC.SG’ can occur. All other forms correspond with those of class 2a. The neuter i-stems were influenced less strongly by other paradigms. The original suffix for the nominative and accusative singular, /-ə/ (< */-i/), has been retained in nouns with short root syllables and was dropped in nouns with long root syllables. The same holds for the nominative/accusative plural suffix /-u/. This results in forms like sper-e ‘spear-NOM/ACC.SG’ and sper-u ‘spear-NOM/ACC.PL (short root) vs. flæ¯sc-∅ ‘flesh-NOM/ACC.SG/PL’ (long root).
1.2.4 Class 4 – Germanic u-stems The nouns of class 4 were either masculine or feminine (but cf. Brunner 1965: Section 275; Hogg and Fulk 2011: Section 2.71 for neuter relics). There is no formal distinction between the two gender values, but there is again a distinction between short and long root syllables. In the former the old stem syllable in the nominative/accusative singular has been retained (cf. Table 18.6), whereas the suffixes for the same values were dropped in the latter (hand-∅ ‘hand-NOM/ACC.SG’). As with most OE nouns, there is some conflation with classes 1 and 2. However the pattern of class 4 has been preserved much better than that of class 3. Table 18.6: Class 4 with short root syllables (sunu ‘son’)
NOM ACC DAT GEN
SG
PL
sun-u sun-u sun-a sun-a
sun-a sun-a sun-um sun-a
1.2.5 Class 5 – Germanic n-stems In Proto-Indo-European, the thematic element of this noun class was formed by a nasal and had the structure /-Vn-/, with the vowel being subject to ablaut alternation. (Here and in the following, I distinguish between “apophony” as a term for any morphological or morphophonemic alternation of a vowel creating an allomorphic distinction irrespective of how it came into being. “Ablaut” only refers to those apophonic alternations that go back to the gradation patterns of Proto-Indo-European; i.e., zero-grade, full grade, etc.) The grammars distinguish three different paradigms for the three gender values, but the patterns differ only slightly in the nominative and accusative singular forms. All other values show the same forms for all three gender values. Cf. Tables 18.7–18.9.
278
III Old English Table 18.7: Class 5 masculine (guma ‘man’)
NOM DAT ACC GEN
SG
PL
gum-a gum-an gum-an gum-an
gum-an gum-um gum-an gum-ena
Table 18.8: Class 5 feminine (tunge ‘tongue’)
NOM DAT ACC GEN
SG
PL
tung-e tung-an tung-an tung-an
tung-an tung-um tung-an tung-ena
Table 18.9: Class 5 neuter (e¯age ‘eye’)
NOM DAT ACC GEN
SG
PL
e¯ag-e e¯ag-an e¯ag-e e¯ag-an
e¯ag-an e¯ag-um e¯ag-an e¯ag-ena
If the root ended with a vowel, the vocalic elements of the suffixes were dropped, as in gefe¯-a ‘joy-NOM.SG’, gefe¯a-na ‘joy-GEN.PL’, gefe¯a-m ‘joy-DAT.PL’ and gefe¯a-n for all other values. Also, some traces of an alternation in the vowel preceding the nasal survived into Old English, as in ox-(e)na ‘ox-GEN.PL’, ox-num ‘ox-DAT.PL’.
1.2.6 Class 6 – Germanic s-stems The original thematic element */-s-/, which, according to Verner’s Law, became */-z-/ and subsequently /-r-/, has been retained in class 6 in the plural forms and occasionally also in the singular where it was lost in most words. Cf. Table 18.10 below.
1.2.7 Class 7 – Germanic r-stems; kinship terms Class 7 consists of only a small number of kinship terms. They had */-r-/ as a thematic element. Except in fæder ‘father’ and sweostor ‘sister’, the dative singular form has an umlauted root vowel. In particular, the masculine nouns of this class show strong influence from class 1, so that the nominative/accusative plural forms fæd(e)r-as can be said to be the common forms in some nouns. Cf. Table 18.11.
18 Old English: Morphology
279
Table 18.10: Class 6 (lamb ‘lamb’)
NOM ACC DAT GEN
SG
PL
lamb lamb lamb-e lamb-es
lamb-ru lamb-ru lamb-rum lamb-ra
Table 18.11: Class 7 (dohtor ‘daughter’)
NOM ACC DAT GEN
SG
PL
dohtor dohtor dehter dohtor
dohtor dohtor dohtrum dohtra
1.2.8 Class 8 – Root nouns The nouns of Class 8 are labelled “athematic” or “root nouns” in traditional descriptions because in earlier stages they lacked a thematic element between root and inflectional suffix. These nouns are either masculine or feminine. The paradigms of the two gender values differ from each other slightly, both because they showed different paradigms already in Proto-Germanic, which later caused different patterns of apophonic alternations, and because they were influenced by the forms of classes 1a and 2 respectively. The nominative and accusative forms of masculine root nouns contained */-i(-)/ in the inflectional suffix in pre-Old English. The respective forms therefore can be identified by the umlauted root vowel (cf. Table 18.12). In northern documents and in early texts, if /-oː-/ is the root vowel, the umlauted vowel is /-œː-/ rather than /-eː-/, represented in the documents by 〈-oe-〉 or occasionally by 〈-œ-〉. The feminine nouns of class 8 originally followed the same pattern as the masculine nouns. However, influence from class 2 was stronger on feminine than that of class 1a on masculine nouns. Particularly in the genitive and dative singular, umlauted root forms competed with non-umlauted affixed forms (cf. Table 18.13). Table 18.12: Class 8: masculine root nouns ( fo¯t ‘foot’)
NOM ACC DAT GEN
SG
PL
fo¯t fo¯t fe¯t fo¯t-es
fe¯t fe¯t fo¯t-um fo¯t-a
280
III Old English Table 18.13: Class 8b: feminine root nouns (bo¯c ‘book’)
NOM ACC DAT GEN
SG
PL
bo¯c bo¯c be¯c, bo¯c be¯c, bo¯c-e
be¯c be¯c bo¯c-um bo¯c-a
1.2.9 Marginal paradigms Some paradigms do not fit in any of the above patterns, either because they are historically derived from extinct noun classes or because several patterns were conflated randomly. Two groups should be mentioned here: Nouns like fre¯ond ‘friend’ or hettend ‘enemy’ are derived from present participles and show the formative element -nd- (cf. below Section 2). However, present participles, if inflected, use the suffixes of the ja-stems (cf. Section 1.2.1), whereas the nouns of this group may follow several different, though not uniform, patterns. Not only is there some inconsistency in the inflectional endings of one word (e.g. frı¯end-∅, fre¯ond-e ‘friend-DAT.SG’; hettend-∅, hettend-e, hettend-as ‘enemy-NOM.PL’), there is also no uniform pattern among these nouns so that there is no point in postulating an independent class for these nouns. Ultimately, these lexemes have to be taken as idiosyncratic. Remainders of the PGrmc. dental stems sometimes still display the dental fricative following the root. Yet, this is again far from resulting in a consistent pattern: ealu exists, for instance, beside ealoþ ‘ale’ and monaþ ‘month’ shows the dental fricative consistently, but its inflection follows that of class 1a. Again, the small number of lexemes to which this applies and the lack of regularity within and across the respective paradigms does not justify postulating an independent noun class for these cases.
1.3 Adjectives OE adjectives agree with the noun they modify in case, number, and gender. For these three categories, there are two different types of declensions of adjectives in Old English. These two patterns are commonly referred to as “strong declension” and “weak declension” and for want of a better term I use these labels here. However, while the same labels are used for different verb classes (Section 2.1), the strong/ weak-distinction of adjectives is functional rather than lexical: the weak declension is used when preceded by a demonstrative or a possessive pronoun, the strong declension is generally used in any other case. In the predicative use, adjectives can be either strong or uninflected. Hogg and Fulk (2011) employ the labels “indefinite” and “definite” adjectives for the two classes, respectively. However, that the pattern of strong and weak adjectives does not quite follow the indefinite/definite distinction, can be seen by the fact that inherently definite pronouns like ælc ‘every’ are followed by a strong adjective. So is the ordinal o¯þer ‘second’ even when it is preceded by a demonstrative. It is therefore difficult to determine a specific function of this distinction since neither [± demonstrative] nor [± definite] reflect the actual use in Old English properly. Cf. the discussion in Mitchell (1985: I, Sections 136–141).
18 Old English: Morphology
281
In poetry, there was a more liberal distribution of the two declensions. If Campbell (1959: 261, Section 638) is right, that “the later the verse the less it diverges from the syntax of prose in this matter”, then this suggests that the distribution between strong and weak adjectives was quite stable up until the end of the OE period and that in an early stage the necessities of the meter could overrule the constraints of adjective inflection. Different from the noun paradigm, there are distinct instrumental forms for strong masculine and neuter singular adjectives. Like in the nominal paradigms, there is some syncretism of forms as can be seen in Tables 18.14 and 18.15 below.
1.3.1 Strong adjectives Strong adjectives follow a pattern which is often referred to as “a- and o¯-declensions”. However, because a number of values are marked by suffixes different from those of noun classes 1 or 2, this label is rather misleading particularly from a synchronic point of view. Those suffixes that do not correspond to the nominal forms are mostly the same as in some pronominal paradigms (cf. below Section 1.5). The default forms for strong adjectives are displayed in Table 18.14. Table 18.14: Strong adjectives (go¯d ‘good’) SG
NOM ACC DAT GEN INSTR
PL
MASC
FEM
NEUT
MASC
FEM
NEUT
go¯d-∅ go¯d-ne go¯d-um go¯d-es go¯d-e
go¯d-e go¯d-e go¯d-re go¯d-re –
go¯d-∅ go¯d-∅ go¯d-um go¯d-es go¯d-e
go¯d-e go¯d-e go¯d-um go¯d-ra –
go¯d-e, go¯d-a go¯d-e, go¯d-a go¯d-um go¯d-ra –
go¯d-∅ go¯d-∅ go¯d-um go¯d-ra –
Some values can differ according to whether the root syllable is long or short. The resulting variation parallels the differences between the noun classes 1a and 1c (Section 1.2.1) for neuter forms and between the noun classes 2a and 2b (Section 1.2.2) for feminine forms. There is also some diachronic variation in the paradigm. Final 〈-e〉 can appear as 〈-æ〉 in early texts. Late texts in West Saxon and Kentish can insert an epenthetic vowel in the endings for the genitive plural (-era) and for the dative and genitive singular of feminine adjectives (-ere). There are a number of phonologically conditioned modifications of the root in some forms of some (groups of) adjectives. These correspond in general with those alternations described in Section 1.2.1 for nouns. Slight variations in the paradigms depending on the existence of a root-final glides /-j-/ and /-w-/ (historically the -ja-/-jo¯- and the -wa-/-wo¯-stems) correspond to those described above for nouns in Section 1.2.1 and in Section 1.2.2.
1.3.2 Weak adjectives The endings of the weak adjectives correspond to those of the nominal class 5 (cf. Section 1.2.5). Only the forms of the genitive plural correspond with the pronominal
282
III Old English paradigms. However, while most forms use the pronominal -ra, some early West Saxon texts attest to the alternative suffix -ena for the genitive plural. In contrast to the strong adjectives, there is no form distinguishing the instrumental case from the dative. Gender distinctions exist only in the nominative and accusative singular. The default forms are shown in Table 18.15: Table 18.15: Weak adjectives (go¯d ‘good’) SG
NOM ACC
PL
MASC
FEM
NEUT
go¯d-a go¯d-an
go¯d-e go¯d-an
go¯d-e go¯d-e
DAT GEN
go¯d-an go¯d-an
go¯d-an go¯d-an go¯d-um go¯d-ra, go¯d-ena (early WSax.)
1.3.3 Comparison of adjectives As in most IE languages, the only independent morphological category of adjectives is comparison with the positive as the default value and the marked values comparative and superlative. The comparative is formed with the suffix -ra. Because this is homophonous with the genitive plural suffix, the forms for this value usually have only one of the two suffixes, so that the form for the comparative in the nominative singular is identical with that for the positive in the genitive plural. An alternative strategy is used in Northumbrian where the suffix -rena is used for the genitive plural of comparative forms. The suffix for the superlative varies between -ost, -ast, and -est. More archaic variants of the same morpheme are -ust (whence -ost). The variation is probably owed to two competing sets of suffixes in Proto-Germanic: *-o¯zan-/*-o¯sta- and *-izan-/*-ista(COMPR/SUP). The fact that some adjectives show i-umlaut in the comparative and superlative forms whereas most adjectives do not (cf. Table 18.16) is further evidence for the different sets of suffixes. The predominance of the superlative -est (< *-ista-) in Old English (and in the later history of English) must therefore have developed after i-umlaut affected the roots. While some adjectives show irregular formations due to phonological cross-influences of suffixes and roots, but use etymologically the same root (labelled “idiosyncratic” in Table 18.16), some frequent adjectives employ hybrid paradigms, i.e., their forms are based on an etymologically different root in the comparative and superlative forms than the positive form. Cf. Table 18.16. Table 18.16: Comparison of adjectives POSITIVE
COMPARATIVE
SUPERLATIVE
regular without i-umlaut
earm ‘poor’ glæd ‘glad’
earm-ra glæd-ra
earm-ost, -ast, -est glæd-ost, -ast, -est
regular with i-umlaut
eald ‘old’ lang ‘long’
ield-ra leng-ra
ield-est leng-est (Continued )
18 Old English: Morphology
283
Table 18.16: Continued
idiosyncratic
hybrid
POSITIVE
COMPARATIVE
SUPERLATIVE
micel ‘great’ ly¯tel ‘little’
ma¯ra læ¯ssa
mæ¯st læ¯st
bet(t)(e)ra se¯lra, sella wiersa
bet(e)st, best se¯lest wierrest, wierst
go¯d ‘good’ yfel ‘bad’
1.3.4 The formation of adverbs There are two (sets of) suffixes to form adverbs from adjectives. Most frequent is the suffix -e, as in clæ¯n ‘pure’, clæ¯n-e ‘purely’. The ultimate source of this suffix is, according to Brunner (1965: 249, Section 315), a PIE ablative marker. Because adverbs from adjectives formed with the adjectivizer -lic frequently occur, the two suffixes (-lic- and -e) are reanalysed as one adverbial marker, so that sometimes adjectives form their adverbial forms with -lice (whence PDE -ly ADV). Moreover, two suffix variants -unga and -inga form adverbs out of any other word class. Hence, a¯n-inga ‘entirely’ from a¯n ‘one’, fæ¯r-inga ‘suddenly’ from fæ¯r ‘attack’, eall-unga ‘entirely’ from eall ‘all’. Some inflected forms of adjectives – mostly genitive forms – can be used adverbially and have been lexicalized as adverbs, e.g. eall-es ‘entirely’ from ‘all-GEN.MASC’, micl-es ‘very’ from ‘great-GEN.MASC’, singal-es ‘always’ from ‘constant-GEN.MASC’. In the same way case/number forms of nouns can be used as adverbial adjuncts from which some forms became lexicalized as adverbs. Examples are dæg-es ‘day-GEN. SG’ > ‘by day’, þonc-es ‘gratitude-GEN.SG’ > ‘willingly’, hwe¯n-e ‘small amount-INSTR. SG’ > ‘a little (ADV)’, fa ¯ cn-e ‘fraud-DAT.SG’ > ‘deceitfully’, hwı¯l-um ‘while-DAT.PL’ > ‘sometimes’. Adverbs formed from adjectives and a number of lexicalized adverbs form their comparative with -or and their superlative with -ost, with some variation in the vowel of the suffix.
1.4 Cardinal Numerals Cardinal numerals are uninflected if they immediately precede a quantified noun. If not, cardinal numerals from ‘1’ to ‘12’ inflect for case and gender. The numerals a¯n ‘1’, and those from ‘4’ to ‘12’ use adjectival endings. The numerals for ‘2’ and ‘3’ have their own paradigms, displayed in Tables 18.17 and 18.18, respectively. They contain partly independent forms, and partly forms that correspond roughly with the pronominal paradigms. Multiples of ‘10’ may or may not be inflected. If they are, the relevant suffixes are those of noun class 1b or of 3. In the genitive they may also use the suffix -ra. The same holds for hund ‘100’ and for þusend ‘1000’. For a detailed description of the Old English numeral system, the numeral forms and their morphological variants, and for the principles of the formation of complex numerals see von Mengden (2010: Chapters II and III).
284
III Old English Table 18.17: Inflectional paradigm of twa¯ ‘2’
NOM/ACC
MASC
FEM
NEUT
twe¯gen
twa¯
twa¯, tu¯
twa¯m, twæ¯m twe¯g(e)a, twe¯g(e)ra
DAT GEN
Table 18.18: Inflectional paradigm of þre¯o ‘3’
NOM/ACC
MASC
FEM
NEUT
þrı¯e
þre¯o
þre¯o
DAT GEN
þrim þre¯ora
1.5 Pronouns The four main types of pronouns in Old English – demonstrative pronouns, personal pronouns, possessive pronouns, and interrogative pronouns – all show slightly different patterns with respect to inflectional categories. When modifying a noun or when used anaphorically, pronouns generally agree with the co-referential noun in number, case, and gender. There is no gender distinction in the plural. If appropriate, differences in the agreement patterns among the four types are mentioned individually in the following sections.
1.5.1 Demonstrative pronouns There are two sets of demonstrative pronouns, often labelled “simple” vs. “composite” (Brunner 1965: Section 337–338: “einfach” vs. “zusammengesetzt”) and “definite article” vs. “demonstrative” (Lass 1994: Section 6.2.2). The latter distinction is justified in so far as the one demonstrative develops into a definite marker. However, although towards the end of the period marking definite reference increasingly becomes its main function, this is never a necessary use and even in late texts definiteness is not systematically encoded. The former distinction – simple vs. composite – is motivated by the etymology of the composite demonstrative, which emerged from the fusion of the inflected form of the simple demonstrative form and a particle *-si. Neither of these sets of labels describes the synchronic situation of the OE period adequately. Because the distribution of the two sets of demonstratives roughly corresponds to that of the PDE pronouns this and that, the terms “distal” and “proximal” appear more adequate. As the distal demonstratives are clearly more frequent, they can be said to be the default set of demonstrative pronouns (cf. Hogg and Fulk 2011: Section 5.3). Both sets of demonstratives distinguish five case/number values, including instrumental forms for masculine and neuter singular. The most frequent forms of the two paradigms are shown in Tables 18.19 and 18.20.
18 Old English: Morphology
285
Table 18.19: Distal demonstratives SG
NOM DAT ACC GEN INSTR
PL
MASC
FEM
NEUT
se þa¯m þone þæs þon, þy¯
se¯o þæ¯re þa¯ þæ¯re –
þæ t þa¯m þæ t þæ s þon, þy¯
þa¯ þa¯m, þæ¯m þa¯ þa¯ra, þæ¯ra –
Table 18.20: Proximal demonstratives SG
NOM ACC DAT GEN INSTR
PL
MASC
FEM
NEUT
þes þisne þissum þisses þy¯s
þe¯os þa¯s þisse þisse –
þis þis þissum þisses þy¯s
þa¯s þa¯s þissum þissa –
1.5.2 Personal pronouns Personal pronouns distinguish case/number as described above. Additionally there are distinct forms for the first, second, and third person. There is no distinction between an inclusive and an exclusive use of the first person plural. The pronoun of the third person also distinguishes gender in the singular. There are four case values (no instrumental) and three number values (singular, dual, plural) for the first and second person and two (no dual) for the third person. The first and second person paradigms show no formal distinction between dative and accusative, except in Anglian texts where the accusative forms mec ‘1P.SG’, uncet ‘1P.DU’, u¯sic ‘1P.PL’ and ðec ‘2P.SG’, incit ‘2P.DU’ and e¯owic ‘2P.PL’ occur. Cf. Tables 18.21–18.23. Table 18.21: Personal pronoun of the first person
NOM ACC/DAT GEN
SG
DU
PL
ic me mı¯n
wit unc uncer
we u¯s u¯re
Table 18.22: Personal pronoun of the second person
NOM ACC/DAT GEN
SG
DU
PL
þu þe þı¯n
git inc incer
ge e¯ow e¯ower
286
III Old English Table 18.23: Personal pronoun of the third person SG
NOM ACC DAT GEN
PL
MASC
FEM
NEUT
he hine him his
he¯o hı¯(e) hire hire
hit hit him his
hı¯(e), he¯o hı¯(e), he¯o him hira, heora
Possessive pronouns are derived from the genitive forms of personal pronouns. In addition to their original case value they can inflect for cases using the suffixes of the strong adjectives (Section 1.3.1). For the third person, there is an independent possessive sı¯n, but it is clearly less frequent than the forms based on his.
1.5.3 Interrogative pronouns Interrogative pronouns have a reduced gender distinction. Nominative and accusative forms distinguish between common and neuter, whereas the other case forms have no case distinction at all. There are forms for five cases, i.e., including the instrumental. Plural forms do not exist. The forms are shown in Table 18.24. Table 18.24: Interrogative pronoun
NOM ACC
MASC/FEM
NEUT
hwa hwone
hwæt hwæt
DAT GEN INSTR
hwa¯m, hwæ¯m hwæs hwı¯, hwy¯
2 Inflectional morphology of verbs 2.1 The morphological categories of verbs All verbs generally encode two tense values, past and present. For both tenses, there are two mood values, indicative and subjunctive. The indicative is the default value and the subjunctive is mainly used when the predication represents the wish of the speaker rather than a real event. In some handbooks it is therefore referred to as “optative”. A third mood value, the imperative, does not distinguish tense. As in all Grmc. languages, there are two main groups of verbs in Old English, traditionally called “weak” and “strong”, which do not differ in the inflectional categories/values they encode, but do differ considerably in the morphological strategy used for encoding the values. While weak verbs employ affixes only (and one circumfix), strong verbs use a system of transfixes (on the distinction between the various types of affixes and the terminology involved cf. Melˈcˇuk 2000: Sections 3.2.2–5; Melˈcˇuk 2006: Sections 3.3.2–5.) The transfix-patterns go back to morphophonemic ablaut alternations in Proto-Indo-European. The affix system of weak verbs emerged during the common Germanic period. Weak and strong verbs will be treated differently in the main sections 2.2 and 2.3 below.
18 Old English: Morphology The verb agrees with the subject in person and number. Number is distinguished between singular and plural (the latter also used when the subject is in the dual). Only in the indicative singular are first, second, and third person distinguished. Generally, the degree of syncretism is considerably smaller among the inflectional markers of verbs than in the nominal paradigms. There are three types of infinite verb forms for each OE verb, the infinitive, a present participle and a past participle, all inflecting for case/number and gender like adjectives (cf. Section 1.3). Analytic constructions with auxiliaries (most commonly beon/wesan ‘be’, habban ‘have’ and weorðan ‘become’) and participles are attested, but they do not seem to be used as systematically for particular tense/aspect/mood values as they are in Present-day English; for details cf. Mitchell (1985: I, Sections 682–743). Finally, the combination of an auxiliary (beon/wesan or weorðan) with the past participle can be used to form past constructions; cf. Mitchell (1985: I, Sections 744–858). On the functions of the two participles cf. further Mitchell (1985: I, Sections 972–989) and on the infinitive cf. Mitchell (1985: I, Sections 920–971).
2.2 Weak verbs 2.2.1 Preliminaries Synchronically, the three classes of weak verbs postulated in traditional descriptions do not differ in the paradigmatic use of inflectional endings. The only differences are morphophonemic alternations in a number of forms affecting either suffix or root which result from different derivational suffixes in Proto-Germanic. Such alternations can, however, also be observed within some of the three classes – resulting from later morphophonemic cross-influences between suffix and root – so that the traditional class distinction is based on an arbitrary choice of criteria. I therefore assume a strictly synchronic perspective and deviate slightly from the common classification. In order to avoid confusion with other descriptions, I use letters instead of Arabic numerals to label the classes. My classes a, b, and c correspond to the three subclasses of what is commonly labelled class 1 (but note that the order of the sub-classes of 1 differs from handbook to handbook), while my classes d and e correspond to the traditional classes 2 and 3, respectively. Of these five classes of weak verbs only class a does not show any significant systematic alternation in the root. I will therefore take this class as the prototypical paradigm, on the basis of which I will discuss some general characteristics of weak verbs and then, in the subsequent subsection, present the other classes.
2.2.2 Class a The default paradigm of weak verbs is exemplified in Table 18.25. It is remarkable that in a strongly inflecting language like Old English (in contrast to both agglutinating and to analytic), there is a suffix (-d-) which is an unambiguous marker of the tense value past. The fact that this suffix -d- is the only morphological marker in all the paradigms of Old English with a truly distinctive one-to-one relation between form and value may indicate that this morpheme is rather young. The marker for the past participle is marked either by the circumfix ge-__-d or by the suffix -d. The two are used in free variation except with prefixed verbs, when the suffix is used throughout.
287
288
III Old English
finite
Table 18.25: Class a of weak verbs (hı¯eran ‘hear’)
SG
P
PRES.IND
IMP
PRES.SUBJ
PAST.IND
PAST.SUBJ
1 2 3
hı¯er-e hı¯er-est hı¯er-þ hı¯er-i-aþ
— hı¯er-∅ — —
hı¯er-e hı¯er-e hı¯er-e hı¯er-en
hı¯er-d-e hı¯er-d-est hı¯er-d-e hı¯er-d-on
hı¯er-d-e hı¯er-d-e hı¯er-d-e hı¯er-d-en
non-finite
PL
INF UNINFLECTED
INFLECTED
PRTC.PRES
PRTC.PAST
hı¯er-i-an
hı¯er-i-enn-
hı¯er-i-end-
(ge-)hı¯er-d-
In many handbooks the suffix -aþ is specified to mark the imperative plural. But since in negations the imperative function can be expressed by the uninflected infinitive or by the subjunctive plural (Brunner 1965: Sections 362.2–3), it seems doubtful to assume a distinct form for the value imperative plural, which is homophonous with the indicative plural and which is not used in a number of contexts which are functionally imperative.
2.2.3 Other classes of weak verbs The following three classes – b, c, and d – show corresponding patterns of a secondary stem formation. In contrast to class a, classes b, c, and d insert a linking element between root and suffix. Synchronically, linkers could be interpreted as a thematic element, but historically they are remainders of derivational suffixes. Within each paradigm, these elements occur in the second and third person of the present indicative, in the imperative, in all past tense forms, and in the past participle. Class b: These are verbs whose root consists of a short syllable ending in /-r-/. Here, the linker is a glide, usually /-j-/, which is represented in the manuscripts either as 〈-g-〉 or as 〈-i-〉. Sometimes, the glide is /-w-/, as in ic gierwe ‘I prepare’, þu gierest ‘thou preparest’. Cf. Table 18.26.
finite
Table 18.26: Class b of weak verbs (herian ‘ravage’)
SG
P
PRES.IND
IMP
PRES.SUBJ
PAST.IND
PAST.SUBJ
1 2 3
her-i-e her-est her-eþ her-i-aþ
— her-e — —
her-i-e her-i-e her-i-e her-i-en
her-ed-e her-ed-est her-ed-e her-ed-on
her-ed-e her-ed-e her-ed-e her-ed-en
non-finite
PL
INF UNINFLECTED
INFLECTED
PRTC.PRES
PRTC.PAST
her-i-an
her-i-enn-
her-i-end-
(ge-)her-ed-
Class c: In principle, the pattern of class c is the same as that of class b. The verbs of this class have short syllables ending in a consonant other than /-r-/. Instead of inserting a glide as linker between root and suffix in the relevant forms, the root-final consonant is geminated. Classes b and c also share the past tense marker -ed- and the imperative singular marker -e and the marker for the past participle (ge-)__-ed. Cf. Table 18.27.
18 Old English: Morphology
289
finite
Table 18.27: Class c of weak verbs ( fremman ‘accomplish’)
SG
P
PRES.IND
IMP
PRES.SUBJ
PAST.IND
PAST.SUBJ
1 2 3
frem-m-e frem-(e)st frem-(e)þ frem-m-aþ
— frem-e — —
frem-m-e frem-m-e frem-m-e frem-m-en
frem-ed-e freme-ed-est freme-ed-e frem-ed-on
freme-d-e freme-d-e freme-d-e freme-d-en
non-finite
PL
INF UNINFLECTED
INFLECTED
PRTC.PRES
PRTC.PAST
frem-m-an
frem-m-enn-
frem-m-end-
(ge-)frem-ed-
Class d: The linking element of this class is /-i-/. In those forms in which this results in a cluster /-i-ə/, an epenthetic glide, spelled 〈-g-〉, is inserted. The markers for the second and third person in the present tense are -ast and -aþ rather than -est and -eþ. The marker for the imperative singular is -a. Also the past tense marker is different from the other classes, i.e., -od, and accordingly, the circumfix for the past participle is ge-__-od. Cf. Table 18.28.
finite
Table 18.28: Class d of weak verbs (lo¯cian ‘look’)
SG
P
PRES.IND
IMP
PRES.SUBJ
PAST.IND
PAST.SUBJ
1 2 3
lo¯c-ig-e lo¯c-ast lo¯c-aþ lo¯c-i-aþ
— lo¯c-a — —
lo¯c-ig-e lo¯c-ig-e lo¯c-ig-e lo¯c-ig-en
lo¯c-od-e lo¯c-od-est lo¯c-od-e lo¯c-od-on
lo¯c-od-e lo¯c-od-e lo¯c-od-e lo¯c-od-en
non-finite
PL
INF UNINFLECTED
INFLECTED
PRTC.PRES
PRTC.PAST
lo¯c-i-an
lo¯c-i-enn-
lo¯c-i-end-
(ge-)lo¯c-od-
Class e: Only four verbs survive that historically go back to a different derivational pattern. They are therefore usually grouped into a distinct class of weak verbs. The relevant verbs are all frequent lexemes (habban ‘have’, libban ‘live’, secgan ‘say’, and hycgan ‘think’) so that their forms are not completely irrelevant in spite of the low type-frequency of this class. There is, as in all marginal classes, considerable variation among the four patterns. It may therefore suffice to say that their inflectional pattern is generally the same as that of the other classes; there is a high degree of alternation in the root because both the root vowel (i-umlaut) and the root final consonants alternate within the paradigms (e.g. habban ‘have’, ic hæbbe ‘I have’, þu hæfst/þu hafast ‘thou hast’).
2.3 Strong verbs Strong verbs use a complex pattern of transfixes to mark tense/aspect/mood, person, and number values. Historically, the paradigm combines an inherited system of suffixation and a systematization of the once morphophonological IE ablaut alternation. For
290
III Old English this reason, the traditional description distinguishes between ablaut vowels and suffixes as markers of strong verbs. Apart from being historically justified, this approach is also motivated by the fact that there are strong parallels in the paradigms of weak and strong verbs in regard to the suffixes encoding person/number distinctions, whereas the apophonic alternation of the root vowels is an exclusive feature of the strong verbs. For each verb paradigm, there are four different variants of root-medial vowels – vowel 1 for the present tense, the infinitive and the present participle: vowel 2 for the first and third person singular of the past tense, vowel 3 for the second person singular and the plural of the past tense and, finally, vowel 4 for the past participle. The seven classes of strong verbs are distinguished by the different sets of root-medial vowels they use in the respective sets of values. Table 18.29 shows the general pattern that holds for all classes of strong verbs. In the following, the four different ablaut vowels are marked by Arabic numerals and are also distinguished by the different shades of the boxes. “C” indicates the two consonantal segments constituting the root template.
finite
Table 18.29: General pattern of transfixes of Old English strong verbs
SG
P
PRES.IND
1 2 3
C-1-C-e C-1-C-st C-1-C-þ C-1-C-aþ
non-finite
PL
IMP
PRES.SUBJ
PAST.IND
C-1-C-∅
C-1-C-e
—
C-1-C-en
C-2-C-∅ C-3-C-e C-2-C-∅ C-3-C-on
PAST.SUBJ
C-3-C-e C-3-C-en
INF UNINFLECTED
INFLECTED
PRTC.PRES
PRTC.PAST
C-1-C-an
C-1-C-enn-
C-1-C-end-
(ge-)C-4-C-en-
Table 18.29 shows that in most cases neither ablaut nor suffix alone suffice to unambiguously encode a particular value, but that both strategies have to be used in combination. From a synchronic perspective, it is in this respect clearly a system of transfixes rather than an ablaut system combined with a set of suffixes (which it is diachronically). But even from a diachronic perspective, the classification is not consistent: the seven sets of apophonic vowels which are taken as definitory of the seven verb classes do not correspond with the original ablaut series of Proto-Indo-European. If the classification were strictly diachronic, classes I to III would have to be comprised as one, similarly classes IV and V. Class VII, by contrast, is based on a completely different morphological strategy, i.e., reduplication rather than apophony (cf. Mailhammer 2007: 53–111). Moreover, the distinction between classes I, II and III is based onthesamegroundsasthesubclassificationofclassIIIintoIIIa,IIIb,andIIIc,namelymorphophonemic alterations. The only period for which the seven classes are postulated to have been regular is Proto-Germanic. However, it remains an open question whether it is a plausible hypothesis to reconstruct a regularsystem, when any earlierand any later stage containedidiosyncrasies. Therefore, although the seven sets of ablaut vowels have some significance for Old English (they do at least have a mnemonic value) they cannot be taken as a morphological system independent of that of the suffixes. I will therefore treat the system of OE strong verbs as a system of transfixes, but retain the traditional categorisation into seven classes. In Table 18.30, the templates of Table 18.29 are applied to the class II verb cre¯opan ‘creep’, so that at least one full paradigm is shown. Table 18.31 presents the vowel series
18 Old English: Morphology
291
of the seven classes and the most important subclasses. Table 18.32 gives one representative of each of the four ablaut vowels for every class. The whole transfixal pattern can then be inferred easily from Table 18.29 and from the respective set of ablaut vowels of Table 18.31. Finally, it should be noted that the suffixes for the second and third person singular indicative caused i-umlaut on the medial vowel (vowel 1 in Tables 18.29 and 18.31). The disruption of the paradigm is sometimes levelled out, but is retained in most cases. The resulting variation of vowel 1 is disregarded in Table 18.29, but can be seen in the present indicative forms of cre¯opan in Table 18.30.
finite
Table 18.30: Class II of Old English strong verbs IMP
PRES.SUBJ
cr-e¯o-p-∅
cr-e¯o-p-e
—
cr-e¯o-p-en
cr-e¯a-p-∅ cr-u-p-e cr-e¯a-p-∅ cr-u-p-on
UNINFLECTED
INFLECTED
PRTC.PRES
PRTC.PAST
cr-e¯o-p-an
cr-e¯o-p-enn-
cr-e¯o-p-end-
(ge-)cr-o-p-en-
SG
P
PRES.IND
1 2 3
cr-e¯o-p-e cr-ı¯e-p-st cr-ı¯e-p-þ cr-e¯o-p-aþ
non-finite
PL
PAST.IND
PAST.SUBJ
cr-u-p-e cr-u-p-en
INF
Table 18.31: The ablaut vowels of the strong verbs
I II IIIa IIIb IIIc IV V VI VIIa VIIb
1
2
3
4
-ı¯-e¯o-i-e-eo-e-e-a-V-V-
-a¯-e¯a-a-ea-ea-æ-æ -o¯-e¯-e¯o-
-i-u-u-u-u-æ¯-æ¯-o¯-e¯-e¯o-
-i-o-u-o-o-o-e-a-V-V-
Table 18.32: Examples of strong verbs
I II IIIa IIIb IIIc IV V VI VIIa VIIb
INF
1P.PAST.IND
PL.PAST.IND
PRTC.PAST
sc-ı¯-n-an cr-e¯o-p-an b-i-nd-an h-e-lp-an w-eo-rp-an b-e-r-an spr-e-c-an f-a-r-an h-a-t-an cn-a¯-w-an
sc-a¯-n cr-e¯a-p b-a-nd h-ea-lp w-ea-rp b-æ-r spr-æ-c f-o¯-r h-e¯-t cn-e¯o-w
sc-i-n-on cr-u-p-on b-u-nd-on h-u-lp-on w-u-rp-on b-æ¯-r-on spr-æ¯-c-on f-o¯-r-on h-e¯-t-on cn-e¯o-w-on
(ge-)sc-i-n-en(ge-)cr-o-p-en(ge-)b-u-nd-en(ge-)h-o-lp-en(ge-)w-o-rp-en(ge-)b-o-r-en(ge-)spr-o-c-en(ge-)f-a-r-en(ge-)h-a-t-en(ge-)cn-a¯-w-en-
‘shine’ ‘creep’ ‘bind’ ‘help’ ‘throw’ ‘bear’ ‘speak’ ‘travel’ ‘command’ ‘know’
292
III Old English The characteristic of class III is that the second root segment consists of two elements. Different consonant combinations in this cluster define the sub-classification of class III: in IIIa, a nasal (N) follows the ablaut vowel (C-V-NC-), in IIIb it is /-l-/ (C-V-lC-) and in IIIc it is /-r-/ (C-V-rC-). Vowel 2 varies in class IIIa due to the free variation between [-a-] and [-ɑ-] of the phoneme /-a-/ before nasals, so that band occurs in the sources equivalent to bond. The second form of class IIIb varies between /-ea-/ in West Saxon and /-a-/ in non-West Saxon varieties. Vowel 3 varies dialectally in classes IV and V between -æ¯- in West-Saxon and -e¯- in non-West-Saxon varieties. In class VI, vowel 1 is to some degree subject to variation, partly dialectal, partly morphophonemic. Synchronically, the verbs of class VII inflect according to the same principle as the other classes of strong verbs. Historically, however, their vowel alternation does not go back to the Indo-European ablaut pattern. Formerly, these verbs formed their past tense through reduplication. The vowel patterns of this class arose through contraction of the root and the reduplicating syllable. In class VII, the root vowels 2 and 3 are always the same. This vowel can be either -e¯- or -e¯o- and the subclassification into VIIa and VIIb is according to this distinction. Vowels 1 and 4 are also the same in almost all verbs of this class. However, there is a greater variety of vowels that are used in these positions. Most frequently it is -a-, but other vowels often occur, while the identity of vowels 1 and 4 is always maintained.
2.4 Irregular verbs Some verbs show patterns that cannot be grouped as weak or strong. Their irregularities have various sources. The most important group of irregular verbs are called “preteritpresent” verbs, because they are derived from past tense forms of (mostly extinct) strong verbs. The past tense acquired a present meaning which then developed an independent inflectional paradigm. For example, the past tense of a verb ‘see’ gives rise to a perfective interpretation. Because ‘having seen’ implies ‘knowledge’, a new verb meaning ‘know’ is lexicalized on the basis of the old past tense form. This is the case with OE wit-an ‘know’ which is cognate with Lt. vid-ere ‘see’. OE witan ‘know’ serves as an example of a paradigm of preterit-present verbs in Table 18.33. However, none of the verbs have identical paradigms and the paradigms of some of these verbs are incomplete, either because the relevant forms are not attested or because they never developed. It is therefore not possible to infer from Table 18.33 the forms of other verbs of this group. Some, but by far not all of them, are used as modal verbs. These are: sceal ‘have to’, mo¯t ‘be allowed to’, mæg ‘be able to’.
finite
Table 18.33: witan ‘know’
SG
nonfinite
PL
P
PRES.IND
PRES.SUBJ
PAST.IND
1 2 3
wa¯t wa¯st wa¯t witan
wite wite wite witen
wiste wiste wiste wistan
INF
PRTC.PRES
PRTC.PAST
witan
witende
gewiten
PAST.SUBJ
– – – –
18 Old English: Morphology
293
There are a few other irregular verbs which are not preterit-present verbs. They share the characteristic that the past tense root differs from that of the present tense. These paradigms can be either idiosyncratic through morphophonemic changes (willan ‘want’ -wold-e ‘want-PAST’; do¯n ‘do’ -dyd-e ‘do-PAST’) or they can be completely hybrid in that there are two completely different roots involved (ga¯n ‘go’ -e¯od-e ‘go-PAST’). Finally, the forms of the copula constitute a hybrid paradigm consisting of three different roots. For some values, forms from two different roots exist. Cf. Table 18.34:
finite
Table 18.34: The copula
SG
P
PRES.IND
IMP
PRES.SUBJ
PAST.IND
PAST.SUBJ
1 2 3
eom/be¯o eart/bist is/biþ sind(on)/be¯oþ
— wes — —
sı¯e/be¯o sı¯e/be¯o sı¯e/be¯o sı¯en/be¯on
w æs wæ¯re w æs wæ¯ron
wæ¯re wæ¯re wæ¯re wæ¯ren
PL
non-finite
BE
INF UNINFLECTED
INFLECTED
PRTC.PRES
PRTC.PAST
wesan/be¯on
wesann-/be¯onn-
wesende/be¯onde
gebe¯on
3 References Blake, Barry J. 2001. Case. 2nd edn. Cambridge: Cambridge University Press. Brunner, Karl. 1965. Altenglische Grammatik: Nach der Angelsa¨chsischen Grammatik von Eduard Sievers. Dritte, neubearbeitete Auflage. Tu¨bingen: Niemeyer. Campbell, Alistair. 1959. Old English Grammar. Oxford: Clarendon. Hogg, Richard M. and R. D. Fulk. 2011. A Grammar of Old English. Vol. II. Morphology. Malden, MA/Oxford: Wiley-Blackwell. Lass, Roger. 1994. Old English: A Historical Linguistic Companion. Cambridge: Cambridge University Press. Mailhammer, Robert. 2007. The Germanic Strong Verbs: Foundation and Development of a New System. Berlin/New York: Mouton de Gruyter. Melˈcˇuk, Igor. 2000. Morphological processes. In: Geert Booij, Christian Lehmann, and Joachim Mugdan in collaboration with Wolfgang Kesselheim and Stavros Skopeteas (eds.), Morphology: An International Handbook on Inflection and Word-Formation, 523–535. Berlin/New York: Mouton de Gruyter. Melˈcˇuk, Igor. 2006. Aspects of the Theory of Morphology. Ed. by David Beck. Berlin/New York: Mouton de Gruyter. von Mengden, Ferdinand. 2010. Cardinal Numerals: Old English from a Cross-Linguistic Perspective. Berlin/New York: Mouton de Gruyter. Mitchell, Bruce. 1985. Old English Syntax. 2 vols. Oxford: Clarendon.
Ferdinand von Mengden, Berlin (Germany)
294
III Old English
19 Old English: Syntax 1 2 3 4 5 6 7
Introduction Word order Noun phrase Verb phrase Complex sentences Summary References
Abstract As a typical Germanic language, Old English has a predominantly synthetic syntax. Flexible word order is determined on pragmatic grounds rather than according to strict syntactic rules. There is a theoretical controversy as to whether the underlying order is SOV or in transition from SOV to SVO. Nouns take four cases, but since in most declensions nominative and accusative forms are identical, the subject-object contrast has to be shown by position rather than by inflectional endings. Finite verbs basically take two tenses (present and preterit), but complex tense forms (especially ancestors of the modern perfect) are also found. The subjunctive mood is the norm in nonfactive contexts and/or in some types of subordinate clauses. Old English makes use of two infinitives (plain vs. inflected) and two participles (active vs. passive). In clause combining paratactic devices (very often repetitive) are much more frequent than subordination.
1 Introduction Old English displays many features of a synthetic language, whose grammar relies heavily on inflectional endings rather than position of sentence elements and use of independent grammatical words. Nouns take four cases (nominative, accusative, genitive, and dative), while adjectives and certain pronouns additionally have the instrumental. Finite verbs have two morphological tenses (present and preterit), but many instances of complex tense-aspect forms, particularly perfect, are also attested. The passive voice is analytic as well. The subjunctive mood has a wide range of predominantly nonfactive uses and is also obligatory in some types of subordinate clauses. Nonfinite forms include the plain infinitive, the inflected infinitive and the active and passive participles. Pragmaticallyoriented word order is far more flexible than in Modern English. Although parataxis is a much more common device of clause combining than subordination, numerous types of subordinate clauses are available. In many cases, however, the borderline between parataxis and hypotaxis is blurred at this early stage of the development of English syntax. There are many parallels between the syntaxes of Old English, Modern German, and Modern Dutch, which is the evidence of their common Germanic heritage. Old English syntax has been studied for more than a century within various theoretical frameworks: traditional philological, structural, generative, semantics-based (including grammaticalization and cognitive theory). We can find purely descriptive accounts and highly abstract theoretical models of explanation. Space does not allow us to present all the interpretations, but some attention will be paid to the main Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 294–313
19 Old English: Syntax theoretical debates. The illustrating language material in this chapter comes mostly from both early and late Old English prose, as poetic texts were subject to the formal requirements of meter and alliteration and also to deliberate archaizing, which affected their syntax significantly. The short titles of the texts follow the standard conventions for Old English (cf. Mitchell 1985).
2 Word order 2.1 Word order patterns As a synthetic language, Old English is characterized by a flexible word order where pragmatic factors such as old vs. new information, topicalization, and the “heaviness” of an element play a decisive role. The categorial status of the element (e.g. pronoun vs. full NP, simple vs. complex VP) matters a great deal, too. There is rich literature concerning Old English word order (e.g. Bean 1983; Mitchell 1985; Traugott 1992; Denison 1993). Within the generative framework (e.g. Traugott 1972; Koopman 1990; Stockwell and Minkova 1991), labels such as SVO, SVO stand for the abstract underlying order rather than the patterns frequently attested in the available texts. Some studies (e.g. Kemenade 1987; Colman 1988; Pintzuk and Kroch 1989) claim that the underlying order was (S)OV, while others (e.g. Vennemann 1974; Allen 1980; Pintzuk 1996) believe that Old English was in transition from the original Germanic (S)OV towards (S)VO (possibly via an intermediate obligatory verb-second stage), especially since later texts have fewer occurrences of verb-final patterns. Finite verb fronting is much more common in main clauses, whereas nonfinites tend to be final. But there is a great deal of variation between OV and VO in all kinds of clauses. All surface arrangements are possible, but statistically subordinate clauses tend to be (S)OV while the most common (and pragmatically unmarked) sequencing in the affirmative main clause is (S)VO, as in (1). When the object is pronominal, it usually precedes the verb (2a), but SVO is also possible (2b): (1)
se cyning 7 þa ricostan men drincað myran meolc (Or 16.7) ‘the king and the noblest men drink mare’s milk’
(2)
a. Gregorius hine afligde (ÆCHom i.22.624) Gregory him expelled ‘Gregory expelled him’ b. Mathathias ofsloh hine sona (ÆLS [Maccabees] 224) ‘Mathathias killed him soon’
The presence of an adverbial at the beginning of a sentence usually triggers inversion, making the clause verb-second XVS(O) (e.g. 3, 4b, 11c, 20a, 30, 50). The same order is also used in main clauses starting with topics (i.e. syntactically prominent elements referring to what the clause is about, as in 5c, 10b) as well as with negative and interrogative words (4a). In dependent questions the subject follows the interrogative pronoun immediately (4b): (3)
Her gewende Cnut cyng to Denmearcon (ChronE 1019) ‘Here (=this year) went King Canute to Denmark’
295
296
III Old English (4)
a. Hwæt sprycest þu cyning? (Bede 196.21) what speakest thou king ‘What do you say, my king?’ b. Đa fregn he mec hwæðer ic wiste hwa ðæt wære (Bede 402.13) then asked he me whether I knew who that were-SUBJ ‘then he asked me if I knew who it was’
Other word orders are possible for pragmatic reasons, such as focus, emphasis, old vs. new information, heaviness of elements, etc. For example, direct objects are given prominence by being fronted at the head of their clauses in (5b) and (5c): (5)
a. Ic ðinum gedwylde dearnunge miltsige (ÆCHom ii.138.2) I thy heresy secretly pity ‘I secretly pity your heresy’ b. Fela Godes wundra we habbað gehyred (ÆCHom i.578.28) many God’s wonders we have heard ‘We have heard of God’s many wonders’ c. hiene ofslog an efor (ChronA 885) him killed one boar ‘a wild boar killed him’ d. ða andswarudon him sume þara bocera (Lk [WSCp] 20.30) ‘then some of the scribes answered him’
SOV
OSV
OVS
VOS
The typical Germanic word order in subordinate clauses (e.g. 6a, 34, 51, 57) and, interestingly, in coordinate main clauses (52, 53), is SOV, but SVO is also possible in the subordinate clause (6b, 43d, 54): (6)
a. … gif he þa ylde hæfde (ÆLS 31.27) if he the age had ‘… if he had been old enough’ b. he cwæð ðæt ðær se iil hæfde his holh (CP 241.5) ‘he said that there the hedgehog had his hole’
Complex verb phrases usually occur in the so-called “brace construction”, where the auxiliary follows the subject immediately and the nonfinite main verb appears at the end of the clause (7a), though the two verbs can also be clustered (7b, 44b, 62). In subordinate clauses the finite auxiliary is usually put at the end, but the brace construction and the juxtaposition can be found too, as in the examples from Orosius under (8): (7)
a. Wif sceal wiþ wer wære gehealdan (Max I 100) wife shall with husband faith hold ‘A wife must keep faith with her man’ b. Se lareow sceal bion on his weorcum healic (CP 81.2) ‘the teacher shall be in his works excellent’
19 Old English: Syntax (8)
a. æfter þæm þe hie þiss gesprecen hæfdon … (51.29) after this that they this promised had ‘After they had promised that …’ þe Philippus hæfde ealle Crecas on b. æfter þæm after this that Philip had all Greeks in his geweald gedon … (118.26) his power put ‘After Philip had put all the Greeks in his power …’ þe Læcedemonie hæfdon oferwunnen Ahtene c. Æfter þæm after this that Laecedemonians had conquered Athens þa burg … (53.20) the town ‘After Laecedemonians had conquered the town of Athens …’
2.2 Subjectless and impersonal constructions Clauses without surface subjects are possible, especially with reference to natural phenomena (9a); cf. Hulk and Kemenade (1993), who discuss the conditions for the “expletive pro-drop” in early Germanic. Apart from such structures, Old English developed the use of the empty pronominal subjects hit and þær, which are neither anaphoric nor cataphoric (9b, c, d): (9)
a. gif on sæternesdæg geðunrað (Prog 1.2 7) if on Saturn’s-day thunders ‘If it thunders on Saturday’ b. hit hagolade seofon niht (Or 123.17) ‘it hailed seven nights’ c. Đa gelamp hit þet se cyng æðelred forðferde (ChronE 1016) then happened it that the king Ethelred died ‘Then it happened that King Ethelred died’ d. þær is mid Estum þeaw (Or 17.6) there is among Ests custom ‘There is a custom among Ests’
Impersonal verbs usually refer to some states of mind, predominantly unpleasant. Since they do not take the nominative subject, the clauses that impersonals appear in are also considered subjectless (cf. Jespersen 1909–49: III, 208–212; Anderson 1986; Denison 1993: Chapter 5). The “logical” subject of the clause denotes the animate experiencer and appears in the oblique case, dative (10a) or accusative (10b), while the cause is optionally expressed by the genitive NP. A significant group of impersonal verbs such as behofian, gedafenian ‘behoove’, (ge)limpan, (ge)weorþan ‘happen’ and þyncan ‘seem’ take finite and/or nonfinite clausal complementation (11a, b). It is remarkable that the verb itself invariably has the third person singular form regardless of the NP form (e.g. 11c); cf. Fischer and van der Leek (1983). Verbs of liking (e.g. cweman, (ge)lician) are often
297
298
III Old English included in the group of impersonals on semantic grounds, though they do have a grammatical nominative subject, the experiencer taking the dative case. The group of impersonal verbs disappeared in Middle English when a reanalysis occurred whereby non-subject NPs were reinterpreted as subjects (Lightfoot 1979: Section 5.1). (10) a. him ofhreow þæs mannes (ÆCHom i.192.16) the man-GEN him-DAT rued ‘He pitied the man’ b. Hine þyrste hwylum 7 hwilum hingrode (WHom 6 168) him-ACC thirsted sometimes and sometimes hungered ‘He was sometimes thirsty and sometimes hungry’ (11) a. ðæm hearpere ða ðuhte ðæt hine nanes ðinges ne lyste (Bo 102.9) thing-GEN not liked that harpist-DAT then seemed that him-ACC no ‘Then it seemed to the harpist that he did not like anything’ b. Đe gedafenað to lerrenne and me to hlistenne (Solil 1 33.4) thee behooves to study and me to listen ‘It is appropriate for you to study and for me to listen’ c. Þonne þuhte eow þas tida beteran (Or 66.1) then seemed-SG you-DAT those times-PL better ‘Then those times would seem better to you’
2.3 Negation The clause is negated by adding the particle ne immediately before the finite verb (12a, 31b). The negative verb is often fronted, as in (13) and (48b). Ne can be reduced (cliticized) and attached to some forms of frequent verbs (wesan, habban, agan, witan, willan, e.g. 12b, 28b, 65). Old English optionally allows multiple negation, otherwise called “negative concord”, as in (13), which is, however, interpreted as a single logical negation. (12) a. He ne andwyrde ðam wife æt fruman (ÆCHom ii.110.33) ‘He didn’t answer the woman at first’ b. Nis angelcynn bedæled drihtnes halgen (ÆLS [Edmund] 259) not-is England deprived Lord’s saints ‘England is not deprived of Lord’s saints’ (13) Ne geseah nan cepa ealand ne weroð, ne geherde non mon not saw no merchant island nor shore not heard no man þa get nanne sciphere, ne furþon ymbe nan gefeoht sprecan (Bo 34.1) then yet none fleet not even about no fight speak ‘No merchant saw an island or a shore; nor did he yet hear anybody speak about any fleet, or even any fight’
19 Old English: Syntax
3 Noun phrase 3.1 Cases The nominative is the (syntactic) subject case, often expressing the agent (14a), and is also used as a form of address (vocative, e.g. 4b, 14b): (14) a. God ‘God b. Sunu son ‘Son,
gesceop us twa eagan and twa earan (ÆGenPref 105) created for us two eyes and two ears’ hwi dydest þu unc ðus? (Lk [WSCp] 2.48) why didst thou us-DUAL thus why did you do it to us?’
The major function of the genitive is to indicate possession, membership, and source. It also has partitive uses, as in (16, 52). The accusative case most commonly expresses the (direct) object, the patient of an action (15). The dative typically refers to the indirect object, the recipient of an action (16). It additionally denotes the experiencer, especially in impersonal constructions (cf. Section 2.2). But case assignment is furthermore an idiosyncratic property of individual verbs, adjectives, and prepositions. There are verbs whose direct objects take the genitive (e.g. fandian ‘try’, brucan ‘use’) and dative (e.g. helpan ‘help’, andswarian ‘answer’). When passivized, these objects retain their case, unlike accusative objects which become nominative under passivization (cf. Fischer et al. 2000: 41), as in (17). (15) ofslogon anne giongne brettisc monnan, swiþe eþelne monnan (they) killed one young British man-ACC very noble man-ACC ‘… slew a young Briton, a very noble man’ (ChronA 501) (16) Þam biscope Wulfhere se cining gesealde landes fiftig hida (LS 3.54) the bishop-DAT Wulfhere the king gave land-GEN fifty hides ‘King Wulfhere gave the bishop fifty hides of land’ (17) him ðurh his hreowsung 7 ðurh Godes miltse geholpen weorðe him-DAT through his repentance and through God’s mercy helped is ‘He is helped by his repentance and God’s mercy’ (CP 251.16) Some verbs (e.g. onfon ‘receive’) govern more than one case without much semantic difference. In the local uses, the accusative usually indicates direction or motion whereas the dative refers to static position, rest. The dative also plays the role of earlier instrumental, ablative and locative, mainly as a complement of prepositions. Moreover, in most declensions nominative and accusative forms are identical, so the subject-object contrast has to be shown by position rather than by inflectional endings. For these reasons the nominative and the accusative are often described in the
299
300
III Old English literature as structural cases (subject vs. direct object) while the genitive and the dative are inherent cases. When morphological contrasts appeared less distinct in late Old English, prepositions became better markers of the earlier semantic case roles.
3.2 Determiners Old English does not yet have the category of the article. Singular indefinite countable nouns normally occur on their own, as in (18), (31). Nonetheless, the numeral an and the indefinite pronoun sum sometimes correspond to the modern indefinite article, introducing new information (19). The demonstrative pronoun is often used as an equivalent of the modern definite article (20a). Unlike in Modern English it can also accompany proper names (20b).
æðelberht cyning in Lundenceastre cirican getimbran (18) Þa heht then ordered Ethelberht king in London church build ‘Then King Ethelberht ordered a church to be built in London’ (Bede 104.21) (19) þin fæder ofsloh an fæt celf (Lk [WSCp] 15.27) ‘thy father slew a/one fat calf’ (20) a. on þysum geare for se micla here þe we gefyrn ymbe in this year went the/that big army that we before about spræcon (ChronA 892) spoke ‘In this year went the big army that we spoke about before’ b. se Cyneheard wæs þæs Sigebryhtes broþur (ChronA 755) ‘that Cyneheard was that Sigebryht’s brother’
3.3 Word order within the
NP
There is a strong tendency for the head to appear at the end of the NP (15, 21), cf. Mitchell (1985: Section 743). Postmodification is mostly found with quantifiers, numerals and -weard words (22), but the same quantifiers can also occur at the beginning of the NP (23). Finally, the determiner and the head can be separated (the so-called floating structures), as in (24). Unlike in Modern English, demonstrative and possessive pronouns can occur in the same NP (25). In appositive structures when one element specifies the other, the name precedes the title, e.g. (3), (16), (18). (21)
seofon slæpera ðrowung (LS 34.1) þæra eadigra the holy-GEN seven sleepers-GEN martyrdom ‘the martyrdom of the holy seven sleepers’
(22) a. þa cingas begen ofslegene wæron (ChronC 868) those kings both slain were ‘Those kings were both slain’
19 Old English: Syntax b. alle Cent eastewearde (ChronA 865) ‘all Kent eastward’ (23)
on eallen Godes beboden (LS 28.33) ‘in all God’s commandments’
(24)
þa comon þa sacerdas to þam cynincge ealle (ÆLS [Book of Kings] 374) then came the priests to the king all ‘Then all the priests came to the king’
(25)
min se leofesta freond (ApT 15.1) ‘my the dearest friend’
There is a tendency to split heavy groups by means of the conjunction and and the extraposition of the conjoined element to the end of the clause, e.g. split objects (26): (26)
he þone cniht agef 7 þæt wif (ChronA 893) he the boy gave and the woman ‘he restored the boy and the lady’
Prepositions usually stand before the NP, but Old English also makes use of postpositions, especially if the NP is a personal pronoun (27a). Postpositions might be a ProtoGermanic heritage and their existence is used as an argument for SOV as the underlying word order in Old English (cf. Kemenade 1987; Pintzuk and Kroch 1989). The postposition is also found in clausal complements (27b), and is obligatory in the relative clauses introduced by þe (27c; also 20a, 57b). Interestingly, no examples of preposition stranding in questions are found in Old English. (27) a. se engel him gewat fram (ÆCHom ii.221.20) the angel them departed from ‘The angel left them’ b. Seo burg wæs swiþe fæger an to locianne (Or 43.23) the town was very fair on to look ‘The town was beautiful to look at’ c. þa adle forecwæde þe heo on forðferde (Bede 318.24) the disease foretold that she on died ‘she foretold the disease of which she died’
3.4 Adjective 3.4.1 Strong-weak distinction Adjectives (and the accompanying pronouns) typically agree in number, gender, and case with the nouns that they modify. A Germanic peculiarity evident in Old English is the morphological distinction of adjectives into strong and weak, which is connected with their definiteness. Typically, strong adjectives modify nouns with indefinite
301
302
III Old English reference (28a), while weak adjectives modify definite nouns (28b). However, some adjectives (quantifiers), e.g. eall, manig, oðer, appear only in strong forms. Predicative adjectives are invariably strong and agree in number and gender with their subject NPs (29). Adjectives tend to occur before the nouns that they modify, which is often used as an argument for the SOV character of Old English syntax (Kemenade 1987; Pintzuk and Kroch 1989). However, when the adjective postmodifies a definite noun, it always takes the strong form (cf. 28b vs. 30). (28) a. Hu mæg he gastlicne wæstm how may he spiritual-ACC.MASC.STRONG fruit ‘How can he have spiritual fruit?’ æþela cyning nolde b. se the noble-NOM.MASC.WEAK king not-would wiðsacan (ÆLS [Edmund] 119) deny ‘The noble king didn’t want to deny Christ’ (29)
Seo eorðe soðlice wæs idel ond æmti (Gen 1.2) ‘the earth truly was desolate and empty’
(30)
on on to to ‘on
habban? (HomS 17 7) have Criste Christ
þone seofoðan dæg eode se cyning sarig the seventh day went the king sorry-NOM.MASC.STRONG þam seaðe (ÆHom 22 484) the pit the seventh day the sad king went to the pit’
3.4.2 Adjective stacking Although Old English usually avoids putting two or three adjectives one after another, preferring to arrange them on both sides of the head or to use coordinate structures, as in (31), the stacking of two and even three adjectives is occasionally used, as in (15) and (32). Both strong and weak forms are attested. However, in later texts we can observe some confusion in the strong-weak distinction, due to the phonetic weakening of inflectional endings. (31) a. Đa tungelwitegan gesawon niwne steorran beorhtne (ÆCHom i.106.24) the astronomers saw new star bright ‘The astronomers saw a new bright star’ b. æþele lareow arfæst 7 gedefe. gesceadwis 7 noble preacher pious and gentle prudent and syfre ne sceolde swa þrowian (ÆCHom i.518.315) temperate not should so suffer ‘a noble pious preacher, so gentle and temperate, should not suffer so much’ (32) a. Þa ageat openlice se earma bearnleasa ceorl (GD 84.51) then found-out openly the poor childless man ‘Then he openly found the poor childless man’
19 Old English: Syntax b. þæt ofstandene þicce slipige horh þu scealt mid the remaining thick slimy humor thou shalt with þam ærgenemnedan læcedomum wyrman (Lch II [2] 16.1.14) the before-mentioned medicine warm ‘You must warm the remaining thick slimy humor with the before mentioned medicine’
4 Verb phrase 4.1 Finite verb 4.1.1 Simple tenses Like other early Germanic languages, Old English has two morphologically simple tenses: present and past (preterit). The present tense refers to the present situation, habitual actions, general states, and the future (33). The preterit refers to past states and events (34). The historical present is not used. Verbs usually agree in number with their subject NPs though inverted word order (e.g. 40b), or conjoining can override that rule. (33) a. Đu wast þæt ic þe lufige (Jn [WSCp] 21.15) Thou knowest that I thee love ‘You know that I love you’ b. ic arise of deaðe on þam þriddan dæge (ÆCHom i.258.7) I will-arise of death on the third day ‘I will be resurrected on the third day’ (34)
sægdon þæt heo swutolice engla song geherdon (Bede 174.14) said that they clearly angels’ song heard ‘(they) said that they clearly heard the angels’ song’
Beon is unique in often having future reference (in addition to expressing habitual states): (35)
He wæs æfre soð Godd 7 is 7 aa bið (WHom 6 138) ‘he was ever true God and is and always will be’
4.1.2 Periphrastic constructions In both poetic and prose texts, we find numerous instances of what look like complex tenses, whose origin and grammaticalization are a matter of dispute among researchers. Lockwood (1968: 114–117) and Traugott (1972: 92) point to many parallels between Germanic and Romance in this respect. Brinton (1988: 99) believes that the development of the perfect in Germanic was preliterary. There are instances of both present and past perfect made up from the auxiliary habban for transitive verbs and the passive participle of the main verb (8, 36) while intransitive verbs typically take the auxiliary beon/wesan (37, 60c):
303
304
III Old English (36) a. Ic hæbbe gebunden þone feond (ÆCHom i.458.18) ‘I have bound the enemy’ b. hig hæfdon heora lufsang gesunggenne (Mt [WSCp] 26.30) they had their hymn sung-ACC.MASC ‘They had sung their hymn’ (37) a. wæs Hæsten þa þær cumen mid his herge (ChronA 894) was Hæsten then there come with his army ‘Then Hæsten had come there with his host’ b. Swæ clæne hio wæs oðfeallenu on Angelcynne (CP Pref 13) so entirely it-FEM was fallen-NOM.FEM in England ‘So completely had it [wisdom] fallen off in England’ The participle is either inflected (in concord with the accusative object of transitives and the subject of intransitives, thus more adjectival), as in (36b) and (37b), or uninflected, especially but by no means exclusively in later texts (Mitchell 1985: Section 710). The loss of inflection indicates reinterpretation toward inclusion of the participle within the VP and the auxiliation of habban/wesan/beon. We can occasionally observe manuscript differences in this respect. Intransitive verbs sometimes also take the auxiliary habban, especially when some kind of accomplishment is described: (38) Þa Moyses hæfde gefaren ofer ða Readan Sæ … (Exod 15.1) ‘When Moses had gone over the Red Sea …’ Especially in later texts the auxiliary and the participle tend to be juxtaposed, as in (8c), (36a), (37b), and (38). Nevertheless, the Old English perfect forms are not fully grammaticalized yet, which is best illustrated by Ælfric’s translation of the Latin plusquamperfectum form steteram by means of ic stod gefyrn (= ‘I stood long-ago’) rather than ic hæfde (ge)standen. The perfect aspect can also be indicated by adverbs, particles, and verbal prefixes (cf. Brinton 1988). Old English also has a construction which is believed to be the ancestor of the modern progressive aspect. However, its etymology and semantics are unclear. Sometimes it resembles Modern English usage (39a), but this “expanded form” (Visser 1963–73: 1920) is also used to describe habitual and permanent states where a modern speaker would use a simple form (39b): (39) a. þæt scip wæs ealne weg yrnende under segle (Or 16.21) that ship Was all way running under sail ‘that ship was running under sail all the way’ b. seo ea bið flowende ofer eal Ægypta land (Or 11.17) that river is flowing over all Egyptians’ land ‘That river flows all over Egypt’
19 Old English: Syntax
4.1.3 Moods The indicative mood refers to facts. There are two non-indicative moods, the imperative and the subjunctive. The imperative expresses orders and prohibitions (40a, b). The personal pronoun is often retained (40b). The first person plural adhortative is formed by means of the auxiliary verb utan, etymologically derived from witan ‘go’ (41). The equivalent of the modern third-person imperative is the hortative subjunctive, as in (42). (40) a. Cedmon, sing me hwæthwugu (Bede 342.26) ‘Caedmon, sing me something’ b. ne nyme ge nan þing on wege (Lk [WSCp] 9.3) not take you-PL no thing on way ‘don’t take anything away’ (41)
utan God lufian (WHom 20.1 121) let-us God love ‘let’s love God’
(42)
Syo hys blod ofer us and ofer ure bearn (Mt [WSCp] 5.23) Be-SUBJ his blood over us and over our children ‘Let his blood be over us and over our children’
The subjunctive is a nonfactive mood, which refers to orders, wishes, fears, doubts, etc., as in (43a, b, c) It expresses improbability and unreality in modally colored contexts and is employed in different types of subordinate clauses (cf. Section 5.3), e.g. (4b), (60b), (65), (68). The subjunctive is also used for indirect discourse (43d, 54a). (43) a. God us gerihtlæce (ÆCHom ii.271.104) ‘May God correct us’ b. ahte ic minra handa geweald (GenA 368) had I my hands power ‘If only I had the power of my hands!’ ne æte (Gen 3.11) c. ic þe bebead þæt þu I thee bade that thou not ate ‘I forbade you to eat’ d. Sume men cweðaþ ðæt hit sy feaxede steorra (ChronA 892) some men say that it be long-haired star ‘Some men say that it is a long-haired star’ Due to the weakening of inflectional endings, the indicative-subjunctive contrast becomes blurred in late Old English, so the morphological subjunctive is being replaced with a periphrastic construction made up of a semantically bleached premodal (especially wolde and sceolde) and the infinitive of the main verb (cf. Traugott 1972; Lightfoot 1979; Warner 1993). First instances are already found in the early Alfredian texts:
305
306
III Old English (44) a. þa Darius geseah þæt he oferwunnen beon wolde þa wolde then Darius saw that he conquered be would then would he hiene selfne on ðæm gefeohte forspillan (Or 70.2) he him self in the fight destroy ‘When Darius saw that he would be conquered, he wanted to kill himself in the battle’ b. ðu geherdest oft reccan on ealdum leasum spellum þætte thou heardst often say in old lying tales that Iob Saturnes sunu sceolde bion se hehsta God (Bo 35.98.25) Jove Saturn’s son should be the highest God ‘You have heard often told in old lying tales that Jove, Saturn’s son, was to be the highest God’
4.2 Nonfinite verb Old English makes use of two kinds of infinitive. The plain (bare) infinitive is the complement of many more verbs than in Modern English, including most preterit-present and causative verbs (45a, b), whereas the inflected infinitive (to V-enne) is the only possible complement of nouns and adjectives (27b) (cf. Callaway 1913). It expresses purpose, obligation (11b), or something imminent, and often has a passive sense (46). There are some controversies concerning its status: some scholars (e.g. Jespersen 1909–1949; Lightfoot 1979) believe to to have been the preposition that governed the dative form of the infinitive (of nominal origin), while others have shown that inflected infinitives behaved more like clauses (e.g. Mitchell 1985: Section 921, Section 3749). The change of the categorial status of the toinfinitive from noun to verb is discussed, among others, by Lightfoot (1979) and Los (1999). The passive infinitive (e.g. 44a) is infrequent and is believed to be a Latin calque. There are instances of what looks like the perfect infinitive (47), but the interpretation of the construction premodal + habban + passive participle is a matter of dispute, owing to the ambiguity of habban, which could have perfect, resultative, experiential and possessive senses. (45) a. Sceap sceal gongan mid his fliese oð midne sumor (LawIne 69) ‘Sheep shall go with its fleece until mid summer’ b. se Cenwalh het atimbran þa ciricean on Wintanceastre (ChronA 643) that Cenwalh ordered build the church in Winchester ‘Cenwalh ordered a church to be built in Winchester’ (46)
ælc broð is to forganne (Lch II 2.23.1.6) ‘every broth is to be avoided’
(47)
for his micclan wundrum þe eft he gedon habban wolde (LS 34.230) for his great wonders that again he done have would ‘for his great miracles that he would work again’
The origin of the accusative with infinitive structure in English is unclear. The structure is probably native with causative and perception verbs (48a), and its use might have been extended to other groups of verbs (of speaking and thinking, e.g. 48b) under the Latin influence. However, Fischer (1989) provides the same syntactic analysis for both kinds. The same verbs are also complemented by participial constructions, as in (49).
19 Old English: Syntax (48) a. he gesawe ðæt leoht of hiofonum on eorðan scinan (Bede 418.11) he saw that light from heaven on earth shine ‘he saw the light shine from heaven to earth’ b. ne tellað we synne weosan gesincipe (Bede 82.4) not consider we sin be wedlock ‘we do not account wedlock a sin’ (49) a. ða sona instepe gefelde ic mec batiende 7 werpende (Bede 404.1) then soon instantly felt I me growing-better and recovering ‘at once I felt myself growing better and recovering’
b. ic wat þæt he mec ofslegene talað (Bede 328.10) I know that he me slain considers ‘I am sure that he accounts me slain’
4.3 Passive voice One of the common ways of expressing impersonality is the use of the pronoun man (grammaticalized from the noun mann) with an active verb form, as in (50). Only one verb hatte (plural hatton) has preserved the Germanic synthetic passive form. Otherwise, the passive voice is periphrastic, made up from the auxiliaries weorðan or beon/ wesan and the passive participle of the main verb, usually in agreement with the nominative subject. There does not seem to be a significant difference in the choice of auxiliaries, though beon/wesan more often refers to the state resulting from some action (51a) rather than to the action itself (17, 51b). (50)
þa sticode him mon þa eagan ut (Or 90.13) then stuck him one the eyes out ‘then his eyes were gouged out’
(51) a. sægde him mon þæt heo of Breotone ealonde brohte said him one that they of Britain island brought wæron (Bede 96.13) were ‘He was told that they had been brought from the island of Britain’ b. Đær wearð Alexander þurhscoten mid anre flan (Or 73.181) ‘There was Alexander pierced with an arrow’
5 Complex sentences 5.1 Parataxis vs. hypotaxis As in other European languages, Latin syntax has provided a pattern for developing Old English complex sentences, which are, however, less hierarchical than their Modern English equivalents and resemble spoken discourse. Apart from asyndetic (uncoordinated) clauses, in the original Old English texts paratactic devices (very often repetitive) are much more frequent than subordination (cf. e.g. Mitchell 1985; Traugott
307
308
III Old English 1992). In fact, the most common conjunction is the word and; e.g., in the Anglo-Saxon Chronicle we find long series of clauses connected just by and. The sequences of events are also expressed by means of clauses beginning with the word þa ‘then’ (e.g. 60a), which develops into a subordinating conjunction equivalent to modern when. Similar ambiguous adverbs/pronouns/conjunctions occur in adjectival and adverbial clauses. Thus the borderline between parataxis and hypotaxis is rather vague at this early stage of development of the English syntax as the syntactic (in)dependence of clauses cannot be easily identified. Even the common multifunctional subordinating particle þe, used on its own or attached to other conjunctions, sometimes appears in coordinate clauses, as well as in main clauses (e.g. 63). The punctuation does not help either, as in most cases it was added by modern editors. Medieval scribes used a completely different system of punctuation to mark pauses for breath rather than syntactic units.
5.2 Coordinate clauses Coordinate clauses often have the “subordinate” word order with a final finite verb. The most frequent coordinating conjunction is and, usually found in manuscripts in an abbreviated form consisting of the Tironian sign 7 (52). The contrasting conjunctions are ac, hwæðere ‘but’, ‘yet’ (53). Other signals of coordination are oððe ‘or’, ne ‘and not’ and correlative conjunctions ge … ge, begen … ge ‘both … and’, ægþer … ge ‘either … or’, nawþer … ne, ne … ne ‘neither … nor’. (52) þa hergodon hie up on Suð Seaxum neah Cisseceastre, then harried they up on South Saxons near Chichester 7 þa burgware hie gefliemdon, 7 hira monig hund and the garrison them drove-away and of-them many hundred ofslogon, 7 hira scipu sumu genamon (ChronA 894) killed and of-their-ships some took ‘Then they harried inland in Sussex near Chichester, but the garrison put them to flight and slew many hundreds of them, capturing some of their ships’ (53) hie ne mehton Suðseaxna lond utan berowan, ac hira they not might South-Saxons’ land out row but of-them þær tu sæ on lond wearp (ChronA 896) there two sea on land threw ‘They were unable to row past Sussex, but there the sea cast two of them ashore’
5.3 Finite subordinate clauses 5.3.1 Nominal (complement) clauses Complement clauses are introduced by the conjunction þæt, sometimes supported by the subordinating particle þe, both merging into þætte (54a, b). The paratactic origin of the structure is evident in the reinterpretation of the demonstrative pronoun as a subordinator. The conjunction can be deleted after some common verba dicendi (55),
19 Old English: Syntax though this is far less frequent than in Modern English. On the other hand, the clause can also be anticipated by the cataphoric pronoun þæt (56): (54) a. Wulfstan sæde þæt he gefore of Hæðum (Or 16.21) ‘Wulfstan said that he went from Hedeby’ b. ongyten þætte þæt is hefig synn (Bede 70.27) ‘(they) understand that that is a grievous sin’ (55)
sægde he he hit gehyrde from þæm seolfan Uttan mæssepreoste (Bede 200.25) said he he it heard from the self Utta mass-priest ‘he said he heard it from the priest Utta himself’
(56)
þa geascade se cyng þæt þæt hie ut on hergað foron (ChronA 911) then learnt the king that that they out on harrying went ‘Then the king learnt that they had gone out harrying’
Many verbs can take both finite and nonfinite complement clauses without any apparent semantic difference, but on the whole nonfinite complementation is less common than in Modern English. Furthermore, there are verbs that take only finite complementation though their modern complements undergo obligatory infinitivization.
5.3.2 Relative (adjectival) clauses Although it is difficult to specify fully reliable rules for the use of relative pronouns, one can observe certain general tendencies, which have numerous exceptions. Restrictive (defining) relative clauses are usually introduced by the indeclinable particle þe, the most frequent relativizer (57a). In late texts we find invariant þæt, which totally replaced þe in Middle English (57b). Less hypotactic nonrestrictive (appositive) clauses tend to have demonstrative pronouns used in the function of relativizers, which have to agree in number, gender and case with the NP that they refer to (58). Both devices can also be combined in restrictive clauses (59). Yet the criteria of the classification are not as clear-cut as in Modern English. (57) a. he ofslog þone aldormon þe him lengest wunode (ChronA 755) he slew the aldorman who him longest remained ‘he killed the aldorman who stood by him the longest’ b. þa halgo rode þet Crist wæs on þrowod (ChronE 963) the holy cross that Christ was on tortured ‘the holy cross on which Christ suffered’ (58)
in Cantwara byrig seo wæs ealles his rices in Canterbury-FEM this-FEM (=which) was all his kingdom’s ealdorburg (Bede 60.11) capital ‘in Canterbury that was capital of all his kingdom’
309
310
III Old English (59)
stodon his geferan oðre þa ðe mid him cwomon (Bede 296.7) stood his companions others those that with him came ‘stood his other companions who had come with him’
5.3.3 Adverbial clauses Old English makes use of the whole range of adverbial clauses, which are usually classified on semantic grounds. Some of them have specific subordinating conjunctions, while others have adopted various pronouns and adverbs to this function (cf. e.g. Kortmann 1997; Lenker and Meurman-Solin 2007). There are both simple conjunctions (e.g. gif, þeah, ær) and complex prepositional phrases playing the role of conjunctions (e.g. forþæm þe, ær þæm þe, foran þa timan þe). The latter type is a peculiar feature of the period, as such formations disappear from English later on. Some clauses obligatorily employ the subjunctive mood. Clauses of place are introduced by conjunctions which were identical in form to spatial adverbs (þær, þider, þanon). There are a number of simple and complex temporal subordinators which denote simultaneity (þa, þa hwile þe), anteriority (ær, ær þæm þe, foran þam (timan) þe, oð), posteriority (siþþan, æfter þæm þe, þæs þe, þa, þonne, sona swa) as in (60): (60)
a. Đa se cyng undergeat þas þing þa ferde he æfter then the king understood those things then rode he after mid þam here (ChronE 1087) with the host ‘When the king understood those things, he went in pursuit with the host’ se hana crawe (Lk [WSCp] 22.61) b. ðu min ætsæcst þriwa todæg ær thou me deniest thrice today before the cock crow-SUBJ ‘You will deny me three times today before the cock crows’ c. þa cwæð se cyningc to his mannum siððan Apollonius then said the king to his men after Apollonius agan wæs (ApT 14.1) gone was ‘Then the king spoke to his men after Apollonius had gone’
Comparative clauses are introduced by (swa) swa, swylce. If the semblance is hypothetical, the subjunctive is used. Clauses of result (61) and purpose are similar, except that the latter use the subjunctive or a premodal, as in (62). The negative purpose is expressed either by negating the lower verb or by the phrase þy læs (þe). (61) þa wæs he sona monad … þæt he wearp þæt sweord onweg (Bede 38.20) then was he soon warned that he threw-IND the sword away ‘then he was soon warned …, so that he threw his sword away’ (62) Þa man sloh eac CC preosta ða comon ðyder þæt hi there one slew also 200 priests who came thither that they scoldon gebiddan for Walena here (ChronA 606) should pray for Welsh-GEN.PL host ‘Two hundred priests were also slain there who had come there to pray for the Welsh host’
19 Old English: Syntax
311
The most common causal conjunction is the complex phrase for þæm/forþon (þe) (literally: ‘for this [that]’), which introduces both subordinate clauses of reason and coordinate clauses of explanation. It is particularly difficult to distinguish between hypotaxis and parataxis as in (63): (63) forþam þe þa iudeas ehton þone hælend forþam þe for-this that the Jews persecuted the saviour for-this that he dyde þas þing on restedaige (Jn [WSCp] 5.16) he did those things on Sabbath ‘The Jews persecuted the Saviour because he did those things on the Sabbath day’ The usual concessive subordinator is þeah (þe). Conditional clauses are introduced either by the conjunctions gif (64), nemne or, less frequently, by inversion (65). Real conditionals use the present tense. Hypothetical clauses employ past verbs, regardless of their time reference. Both clauses of the unreal conditional period make use of parallel past tense forms (65, 68), but in later and northern texts we find the first uses of periphrastic structures with a premodal in the main clause (66b): (64)
Gif þu hunig to dest þæt deah (Lch II 2.11.5) If thou honey to dost that avails ‘If you add honey, it is good’
(65)
swa manega martyras nære seo mycele ehtnyss (ÆLS 11.328) Næron not-were so many martyrs not-were the big persecution ‘There would not have been so many martyrs had there not been this great persecution’
(66) a. gif god wære eowre fæder. witodlice ge lufedon me (Jn [WSCp] 8.42) b. gif god faeder iuer uoere gie ualde lufiga uutudlice mec (JnGl [Li]) ‘if God were your father you would truly love me’
5.3.4 Correlative adverbs Subordinating conjunctions of the lower clause are often accompanied by correlative adverbs in the main clause. In numerous cases the adverb is identical with the subordinator (þa … þa …, forþon … forþon …), which often makes it difficult to determine which of the clauses is subordinate. The presence of the subordinating particle þe sometimes helps (67), but the problem is that it can also occur in the main clause. The situation is much clearer in conditional and concessive clauses, where the correlative adverb is a different word (68). (67) Þider ðe Stephanus forestop … þider folgode Paulus (ÆCHom i.202.116) thither that Stephen stepped thither followed Paul ‘Where Stephen stepped … Paul followed him’ (68) gif If
þu wistest hwæt þe thou knewest what thee
toweard future
is is
þonne then
weope wept-SUBJ
312
III Old English þu mid me (ÆCHom i.412.67) thou with me ‘If you knew what is to come to you, then you would weep with me’
6 Summary Old English syntax was typically Germanic, in that synthetic devices prevailed. In the following centuries the English language became more analytic due to the reduction of inflections brought about by the weakening of final syllables, accelerated by extensive foreign contacts. This resulted in the fixing of the SVO word order and rapid development of periphrastic constructions (e.g. complex verb phrases making use of auxiliaries, prepositional phrases). Furthermore, Old English paratactic ways of combining clauses were gradually replaced with more and more common as well as more complex and hierarchical subordination, including the use of nonfinite constructions.
7 References Allen, Cynthia. 1980. Movement and deletion in Old English. Linguistic Inquiry 11: 261–323. Anderson, John. 1986. A note on Old English impersonals. Journal of Linguistics 22: 167–177. Bean, Marian. 1983. The Development of Word Order Patterns in Old English. London: Croom Helm. Brinton, Laurel J. 1988. The Development of English Aspectual Systems. Cambridge: Cambridge University Press. (Reprint: 2009) Callaway, Morgan. 1913. The Infinitive in Anglo-Saxon. Washington, DC: Carnegie Institution. Colman, Fran. 1988. Heavy arguments in Old English. In: John Anderson and Norman MacLeod (eds.), Edinburgh Studies in the English Language 1: 33–89. Denison, David. 1993. English Historical Syntax: Verbal Constructions. London: Longman. Fischer, Olga. 1989. The origin and spread of the accusative and infinitive construction in English. Folia Linguistica Historica 8: 143–217. Fischer, Olga, Ans van Kemenade, Willem Koopman, and Wim van der Wurff. 2000. The Syntax of Early English. Cambridge: Cambridge University Press. Fischer, Olga and Frederike van der Leek. 1983. The demise of the Old English impersonal construction. Journal of Linguistics 19: 337–368. Hulk, Aafke, and Ans van Kemenade. 1993. Subjects, nominative case, agreement and functional heads. Lingua 89: 181–215. Jespersen, Otto. 1909–49. A Modern English Grammar on Historical Principles. 7 vols. London: George Allen and Unwin. Kastovsky, Dieter (ed.). 1991. Historical English Syntax. Berlin/New York: Mouton de Gruyter. Kemenade, Ans van. 1987. Syntactic Case and Morphological Case in the History of English. Dordrecht: Foris. Koopman, Willem. 1990. Word Order in Old English: with Special Reference to the Verb Phrase. Amsterdam: University of Amsterdam. Kortmann, Bernd. 1997. Adverbial Subordination. A Typology and History of Adverbial Subordinators Based on European Languages. Berlin/New York: Mouton de Gruyter. Lenker, Ursula and Anneli Meurman-Solin (eds.). 2007. Connectives in the History of English. Amsterdam/Philadelphia: John Benjamins. Lightfoot, David. 1979. Principles of Diachronic Syntax. Cambridge: Cambridge University Press. Lockwood, William Burley. 1968. Historical German Syntax. Oxford: Clarendon Press. Los, Bettelou. 1999. Infinitival Complementation in Old and Middle English. The Hague: Holland Academic Graphics. Mitchell, Bruce. 1985. Old English Syntax. 2 vols. Oxford: Clarendon Press.
20 Old English: Semantics and lexicon
313
Pintzuk, Susan. 1996. Old English verb-complement word order and the change from OV to VO. York Papers in Linguistics 17: 241–264. Pintzuk, Susan and Anthony Kroch. 1989. The rightward movement of complements and adjuncts in the Old English of Beowulf. Language Variation and Change 1: 115–143. Robert, Stockwell and Donka Minkova. 1991. Subordination and word order change in the history of English. In: Kastovsky (ed.), 367–408. Traugott, Elizabeth Closs. 1972. A History of English Syntax. New York: Holt, Rinehart and Winston. Traugott, Elizabeth Closs. 1992. Syntax. In: Richard M. Hogg (ed.), The Cambridge History of the English Language. Vol. I. The Beginning to 1066, 168–289. Cambridge: Cambridge University Press. Vennemann, Theo. 1974. Topics, subjects, and word order: from SXV to SVX via TVX. In: John Anderson and Charles Jones (eds.), Historical Linguistics I: Syntax, Morphology, Internal and Comparative Reconstruction. Amsterdam: North-Holland, 339–376. Visser, Frederikus Theodor. 1963–73. An Historical Syntax of the English Language. 3 vols. Leiden: E. J. Brill. Warner, Anthony. 1993. English Auxiliaries: Structure and History. Cambridge: Cambridge University Press.
Rafał Molencki, Katowice (Poland)
20 Old English: Semantics and lexicon 1 2 3 4 5 6 7
Introduction Resources The nature of the lexicon Innovation and change The nature of the evidence Word and field studies References
Abstract Following an introduction on the relationship between Old and Modern English vocabulary, Section 2 outlines resources for studying the OE lexicon, notably the Dictionary of Old English (DOE) (Cameron et al. [eds.] 1986–) at the University of Toronto; the online third edition of the Oxford English Dictionary (OED) (Simpson [ed.] 2000–); and the Historical Thesaurus of the Oxford English Dictionary (HTOED) (Kay et al. 2009). Section 3 deals with the nature of the lexicon, covering the size of the vocabulary, cognate words, foreign language loans, and features of lexical structure such as affixation and compounding. Section 4 is concerned with aspects of innovation and change, particularly polysemy and homonymy as processes of semantic change and cognitive approaches to metonymy and metaphor. The concluding sections deal with some of the problems posed by the nature of the evidence in studying the OE lexicon, and interdisciplinary word and field studies in a selection of projects.
Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 313–325
314
III Old English
1 Introduction The vocabulary of Old English forms the core lexicon of the many varieties of English spoken around the world today. Even without a knowledge of Old English, it is not difficult to see the connection between such OE and modern pairs as eorðe > earth, sæ¯ > sea, mo¯dor > mother, fo¯t > foot, go¯d > good. Examples such as these abound in shared areas of vocabulary like those dealing with the natural world, people, relationships, and activities. Common grammatical words, such as the conjunction and or the preposition in, also survive, though their range of meaning may not be identical to that in Modern English. On the other hand, many OE words would be unfamiliar to a modern reader, either because they have disappeared from use or because they designate objects, concepts or activities which are no longer part of people’s lives. According to Baugh and Cable (1993: 53), only around 15% of recorded OE words have survived into Modern English. Others again may never have reached us, since our knowledge of Old English depends on the hazardous transmission of texts over a period of more than a thousand years. Such texts as survive, moreover, offer limited evidence since they mostly represent formal and literary registers rather than everyday spoken language.
2 Resources Resources for the study of Old English are continually improving (Lowe, Volume 2, Chapter 71; Traxel, Volume 2, Chapter 72; Durkin, Volume 2, Chapter 73). For students of the OE lexicon, by far the most significant development in recent years has been the ongoing Dictionary of Old English (Cameron et al. [eds.] 1986–) project at the University of Toronto. Using a computerized corpus containing at least one copy of every surviving OE text, the team of lexicographers is engaged in a comprehensive reexamination of the entire OE lexicon, employing contextual analysis to determine and group meanings. DOE is thus a huge step forward in what up until now has been a largely 19th century dictionary tradition. Special features which are particularly useful for lexicologists are those indicating the currency of a word, such as information about frequency of occurrence and the inclusion of usage labels where there is significant restriction to a particular period, dialect, author or genre. Spelling variants are also listed. Publication is by fascicle in a variety of formats, including DOE: A to G online (Cameron et al. 2007), which offers Boolean searches on the ten fields in which the various types of information are stored and has the added bonus of providing links to the online Oxford English Dictionary (Simpson [ed.] 2000–), thereby allowing the user to trace the development (or absence) of OE words in later periods. The corpus on which the dictionary is based, comprising over 3,000 texts, has also been released, enabling scholars to see for the first time the full range of contexts in which a word appears. A further step forward is the substantial revision of Old English materials in the third edition of the OED, although there has been no change in that dictionary’s policy of excluding words which became obsolete before 1150. Esposito (2002) describes the scope of the revision: Every single Old English quotation, whether already in OED or newly added, is being checked against the most recent reliable edition of the text, with new bibliographical
20 Old English: Semantics and lexicon details and additional context being given where appropriate. Dating of quotations has been radically revised, with NED’s [New English Dictionary] assumed composition dates replaced by a simple threefold division of all pre-1150 quotations into “early OE” (up to 950), “OE” (950–1100), and “late OE” (1100–1150), based firmly on manuscript dates as agreed by the most recent scholarship.
An onomasiological version of the OED has been produced at Glasgow University and published as The Historical Thesaurus of the Oxford English Dictionary (HTOED) (Kay et al. 2009), enabling study of the development of meaning from Old English to the present within a framework of semantic categories. The coverage of Old English in HTOED has been extended beyond the OED’s provision by integrating materials from A Thesaurus of Old English (TOE) (Roberts and Kay 2000) into this framework and joining them where appropriate to OED entries. HTOED is based on the second edition of the OED (Simpson and Weiner 1989), and dates all Old English words simply as “OE”, but will be linked to the OED as revision progresses.
3 The nature of the lexicon Figures derived from TOE (Roberts and Kay 2000) give a total of around 34,000 separate word forms in Old English, less than half the number that might be found in a modern desk dictionary. The total rises to 50,700 meanings if polysemy and the occasional case of homonymy are taken into account. For comparison, DOE: A to G online (Cameron et al. 2007), which covers the first eight of the 22 letters of the OE alphabet, contains 12,568 headwords. In TOE, nouns predominate at just over 50%, followed by verbs at 24% and adjectives at 19%. The OE figures will undoubtedly change as editing of DOE progresses (see Section 2 above). Any examination of the OE lexicon reveals its essentially Germanic character. Words like those in Section 1 often have cognate forms in other Germanic languages, for example modern German Erde, See, Mutter, Fuss, gut, or Swedish jord, sjo¨, moder, fot, god. The differences between cognate languages, and the differences between old and modern versions of the same language, show how word forms develop and diverge over the years. Compared with Modern English, Old English contains very few words borrowed from foreign languages. When the Anglo-Saxons arrived in Britain, their language already contained some words borrowed from Latin through contact with Roman activities on the European mainland. These include coper ‘copper’, stræ¯t ‘road’, and wı¯n ‘wine’. Following the conversion of the Anglo-Saxons to Christianity, Latin terms increasingly appear in the vocabulary of religion and education as well as in more general areas where new commodities, ideas or practices were introduced. From the several hundred words recorded, examples include abbod ‘abbot’, sealm ‘psalm’, sco¯l ‘school’, discipul ‘disciple, student’, plante ‘plant’. Many individual plant-names, often for plants useful in medicine, were borrowed from Latin. Religious influences also came from France, and a few French loans are recorded late in the OE period, notably pru¯d ‘proud, arrogant’, leading to derived forms such as oferpru¯t ‘haughty’ and woruldpru¯do ‘worldly pride’. Native words, however, might continue to be preferred over synonymous foreign ones. Discipul was a relatively rare word in OE; the word used in the Anglo-Saxon Gospels and elsewhere was the native leorningcniht. They might also be
315
316
III Old English more productive: unlike plante, native wyrt ‘plant, herb’ generates a host of compounds, such as wyrtcynn ‘species of plant’. (See further, Wieland, Chapter 23.) A mere handful of words, perhaps around 20 in all (Hogg 1992: 3), were borrowed into the general language from the Celtic-speaking people who already inhabited Britain. The best known of these are probably brocc ‘brock, badger’ and a¯ncor ‘anchorite, hermit’. According to Breeze, however, many Celtic loans in English remain to be discovered: he puts forward a case for, among others, OE syrce ‘coat of mail’ and trum ‘strong’ (Breeze 2002: 175–176). Less controversial is the fact that many place-names in certain parts of the British Isles are Celtic in origin (see also Hough, Chapter 14). A more significant contact, linguistically at least, was with the Old Norse (ON) language of the Scandinavian Vikings, who raided, and later settled in, much of the east and north of the country. Unusually, and probably because of the cognate nature of the two languages and the fact that transmission occurred during everyday spoken interaction, Scandinavian-derived words replaced their OE counterparts in core areas of the language, resulting in Modern English words such as take (OE (ge)niman), sky (OE lyft) and the pronoun they (OE hı¯e). Often the cognate words were very similar in form, as OE sweostor and ON syster, the latter giving Modern English sister. Because such words were likely to have been restricted to casual spoken use in the early stages, only a few of them appear in the OE written record, but many more are found in early Middle English. Thus, take (OE tacan) is recorded in the OED late in the OE period, but sky is not listed until the 13th century, although it was probably in use before then (see Dance, Volume 2, Chapter 110). A full account of foreign borrowings into Old English is given in Baugh and Cable (1993: 72–104) and Kastovsky (1992: 299–338). Words throughout this paper are generally given in the form found in Clark Hall’s A Concise Anglo-Saxon Dictionary (1960); Clark Hall’s brief definitions are also followed. A useful feature of this dictionary for those interested in tracing the development of words is the inclusion of OED headword forms of OE words surviving into Middle and Modern English.
3.1 Lexical structure: affixation Basic OE words tended to be short forms of one or two syllables. Stress fell on the root syllable, which was usually the first syllable. Grammatical information was conveyed by variable endings on words, identifying their role in the clause (see von Mengden, Chapter 18). Prefixes and suffixes were added to roots to create a variety of kinds of new words. In general, prefixes tended to change meaning, for example by negating or intensifying the root meaning, as in oferfull ‘too full’ or mislæ¯dan ‘mislead’. Prefixes were often used to form verbs, for example u¯pfle¯ogan ‘to fly up’ and a¯fle¯ogan ‘to fly away, flee’ from fle¯ogan ‘to fly’. Suffixes were used to create different parts of speech, such as the adverb hearde ‘fiercely’ from the adjective heard ‘hard, fierce’. Many OE adjectives end in -ful (caru/cearu ‘care, sorrow’, carful/cearful ‘sorrowful’), -ig (wæter ‘water’, wæterig ‘watery’), -isc (cild ‘child’, cildisc ‘childish’), -le¯as (lı¯f ‘life’, lı¯fle¯as ‘lifeless’), -lic (sige ‘victory’, sigelic ‘victorious’). Common adverbial suffixes include -e (de¯op ‘deep’, de¯ope ‘deeply’) and -lı¯ce (de¯oplic ‘deep’, de¯oplı¯ce ‘deeply’). Both -end and -ere were used to form agent nouns, as in lærend ‘teacher’ and leornere ‘pupil, disciple’. Abstract nouns often end in -do¯m (wı¯s ‘wise’, wı¯sdo¯m ‘wisdom’), -ha¯d (cild ‘child’, cildha¯d
20 Old English: Semantics and lexicon ‘childhood’), -nes (yfel ‘evil’, yfelnes ‘wickedness’), -scipe ( fre¯ond ‘friend’, fre¯ondscipe ‘friendship’). Other common Modern English suffixes, such as those in words like emotion, magnitude, generous, generosity, social, sociable, sociability, were adopted after the OE period from French or Latin. One result of the frequency and flexibility of word formation in Old English is that we often find groups of words clustered round a shared root, as in the following words derived from sorg ‘sorrow, distress’: sorgung ‘sorrowing’, sorgful ‘sorrowful’, sorgle¯as ‘sorrowless’, sorig ‘sorry’, sorgian ‘to feel sorrow’, unsorh ‘unsorry, free from care’. All of these affixes, except for the -an which indicates the infinitive form of the verb in sorgian, survive in Modern English, although particular forms and meanings may have been lost. For example, an adverb from the group, sorglı¯ce ‘miserably’, survives into Middle English as sorrowly, with a last date in the 13th century, but of the adjective sorglic ‘miserable’, which might have survived in the same form, there is no trace beyond Old English. Likewise, there is no trace in the record of unsorh between Old English and the 20th century, where the OED (s.v. un- prefix1, def. 7) finds three citations for unsorry. This may be an accident of collection, or may reflect the flexibility of prefixes such as un-, which speakers can use to invent new words as occasion demands. Sometimes prefixes have little if any effect; giefan and forgiefan, for example, both mean ‘to give’, although only forgiefan develops the meaning ‘forgive’. Many verbs may occur with or without the prefix ge-: niman and geniman both mean ‘to take’. Such variation is sometimes summarized in OE dictionaries and grammars by bracketing the prefix, as in (ge)niman, and the ge is ignored for purposes of alphabetization.
3.2 Compounds The root sorg also yields a number of characteristic OE compounds, where two independent words are joined to express a complex idea, as in sorg plus cearu ‘care’, yielding sorgcearu, meaning ‘anxiety’. Compounding was a favourite way of creating new words in Old English, with the combination of two nouns, as in sorgcearu, being the most frequent type (Kastovsky 1992: 365). Other types include noun plus adjective (nihtlang ‘night-long’), adjective plus adjective (blæ¯hæ¯wen ‘light blue’) and adjective plus noun (ealdfæder ‘forefather’). However, as Hogg (1992: 23–24) points out, we often cannot be sure from a manuscript, let alone a subsequent edition, whether one word or two was intended; possible solutions to this problem are discussed by Kastovsky (1992: 362–363), although it may be a problem which bothers modern readers, used to the consistent conventions of the printed page, more than it did Anglo-Saxon scribes. Compounds were used where Modern English is more likely to use a phrase, as in sorglufu ‘sorry or sad love’. Sometimes they contained a good deal of information, as in heorotsol ‘a stag’s wallowing place’ or paddanı¯eg ‘an island populated by toads or frogs’. Many of them have disappeared from the language: we no longer express distress with sorgword ‘sad words’ or sorgle¯oþ ‘sad song’, but with the Latin-derived dirge or with lamentation, also from Latin but possibly entering English through French; these are first recorded in the OED in c.1225 and 1382 respectively (s.v.v. dirge n. and lamentation n.). The group centring on sorg thus illustrates in microcosm both how the OE vocabulary was structured and how the language has changed since OE times.
317
318
III Old English Many compounds, such as those above, are transparent in meaning, i.e. the meaning of the whole is obvious from its parts. Others, known as “kennings”, are more opaque, relying on a metaphorical interpretation. Kennings, and compounds generally, abound in OE poetry and therefore refer to subjects often treated in poetry, such as emotions, epic voyages, and heroic deeds. Thus we find kennings for the sea like swanra¯d ‘swan’s road’, hwælweg ‘whale’s path’, and fisces bæð ‘fish’s bathing place’ (which may be a phrase rather than a compound). If we look up expressions for ‘ship, boat’ in A Thesaurus of Old English (Roberts and Kay 2000: 331), we find 47 general words for the concept as well as 42 more specialized ones. Such a high degree of lexicalization, comprising both synonyms for the central concept and words for more specific concepts associated with it, indicates the importance of this concept in the culture of the time. Many of these words occur in poetry, often only in poetry. By far the most frequent metaphor is that of the horse, a common mode of transport on land at the time, shown in examples such as brimhengest, merehengest, sæ¯mearh, sundhengest and y¯ðmearh, where the first element means ‘sea’ and the second ‘horse’. Some of these compounds also occur in more prosaic contexts; for example sæ¯genga, meaning ‘sea companion, ship’ in the poem Beowulf, is used more literally elsewhere to mean ‘sailor’, while sæ¯hengest means ‘hippopotamus’ as well as ‘ship’. A vexed, and probably unanswerable, question about such words, as about synonyms generally, is whether an Anglo-Saxon speaker would be aware of their etymological differences and possible shades of meaning or would simply regard them as approximately synonymous and thus interchangeable in most contexts. Taken together with the frequent repetition of initial sounds, these examples also reflect the twin demands of Old English poetic style, alliterating stressed syllables and “elegant variation” through synonymy (see Fulk, Chapter 25). A comprehensive treatment of all aspects of word-formation can be found in Kastovsky (1992: 355–400) and the works cited there.
4 Innovation and change All languages have ways of acquiring new words as the need arises. As we have seen in Section 2, Old English, like other Germanic languages past and present, favored using internal resources such as affixation and compounding for this purpose, but occasionally borrowed words from foreign languages. Since Old English was a predominantly synthetic language, using inflectional endings to express grammatical relationships, words could not usually be borrowed in the foreign form but had to be adapted to fit OE patterns, as when the Latin word discipulus ‘a disciple’ was adopted into Old English as discipul. Sometimes words from two sources existed side by side for a time. For example, alongside discipul we find native derivatives such as leornere ‘learner’ and compounds such as leorningcniht and leornungmann ‘learningboy/man’, the latter glossed in Clark Hall (1960: 216) as “used even of women”. Sometimes the foreign word is effectively translated into Old English, reproducing the form of the loanword in what is termed a “loan-translation” or “calque”. Thus the Latin word patriarcha ‘chief father/bishop, patriarch’ becomes OE he¯ahfæder ‘high father’. By a similar process, Latin sanctus ‘holy person, saint’ becomes OE ha¯lga ‘holy one, saint’, and trinitas ‘group of three, Trinity’ becomes ðrines ‘threeness, Trinity’. It is typical of the history of English vocabulary that the OE terms were replaced in later periods by borrowing
20 Old English: Semantics and lexicon the Latin words which they had once translated. However, we do retain the expression Holy Ghost, OE Ha¯lig Ga¯st, a calque of Latin Spiritus Sanctus, rendered somewhat strange to modern ears by the narrowing of meaning of the word ga¯st to refer to the particular kind of spirit we call a ghost. One of the commonest, most economical (and least noticeable) ways of supplying a new word at all periods of English is to extend the meaning of an existing one, for example to embrace a new concept. Following the introduction of Christianity, concepts such as ‘God’, ‘heaven’ and ‘hell’ took on new meanings for the Anglo-Saxons but were expressed by words which had referred to similar concepts in the old religion: God, heofon, hell. The use of such familiar terms presumably made the new ideas more acceptable to potential converts, and illustrate the effect that cultural change can have on language.
4.1 Polysemy and homonymy Various processes of semantic change can bring about the condition known as polysemy, where a single form has two or more distinct but ultimately related meanings either simultaneously or at different stages of a language’s development. Sometimes a borrowed word already has more than one meaning, as in torr ‘rock, crag’ and ‘tower, watch-tower’, which had both meanings in the original Latin before the word entered Old English through Celtic. Two of the commonest of these processes of change are “narrowing” or “specialization”, where a word’s meaning becomes more restricted, and “pejoration”, where the word comes to refer to something which is regarded as in some way inferior. For example, the OE word fe¯ond meant both an enemy and, by a process of narrowing, the supreme enemy, the Devil. Likewise, the word æppel in Old English usually referred to any kind of fruit, as in palmæppel ‘fruit of the palm, date’, but there is evidence in the OE corpus of the beginning of a narrowing process to meaning the fruit we now call an apple. Narrowing often precedes pejoration. The word cniht basically meant ‘a boy, youth’, but came to refer to those performing roles commonly filled by boys, including the role of servant. In this case, the role was often at the relatively high social level of an attendant or retainer, resulting eventually in the modern word knight. However, in the case of cnafa/cnapa, meaning ‘child, youth’ and then ‘servant’, pejoration gave us ModE knave. The parallel processes of widening or generalization and amelioration are much rarer. An example of the former can be found in the word hla¯ford ‘lord’, which originally referred to the specific role of a lord within the Anglo-Saxon social system, but was extended more generally to people in authority, leading to compounds such as hla¯forddo¯m and hla¯fordscipe, both meaning ‘authority’. This word also exhibits narrowing in its meaning of ‘husband’ and possibly amelioration as one of the many OE terms for ‘ruler’ applied to the Christian God. Most semanticists distinguish between “polysemy”, where new meanings are linked to old, and “homonymy”, where two words just happen to have the same form through historical accident. There are very few homonyms in Old English, both because its vocabulary derives largely from a single source and because it is an inflected language, less hospitable to borrowed forms, which are a frequent cause of homonymy. DOE treats as homonyms the etymologically unrelated fa¯h ‘at feud, hostile’ and fa¯h ‘variegated, stained, shining’, but Healey (2006: 85–86) notes that the distinction is not always clearcut and can be deliberately exploited to create ambiguity. Possible occurrences of homonyms are often masked by the fact that modern editions of OE texts indicate
319
320
III Old English vowel length by a diacritic; Anglo-Saxon scribes did not use such marks. Thus ac ‘but’ and a¯c ‘oak’ would have looked the same on the manuscript page, as would sæl ‘room, hall’ and sæ¯l ‘time, season’ or broc ‘misery, affliction’ and bro¯c ‘brook’. Homonymy can cause ambiguity in understanding a text if both words make sense in a given context, which seems unlikely in these cases. However, much critical ink has been spilled over the interpretation of gæst in line 2312 of the poem Beowulf, describing the first appearance of the dragon that will eventually kill the hero: (1)
Đa se gæst ongan gle¯dum spı¯wan. Then the ? began fire to spew forth ‘Then the ? began to spew forth fire’
Is our mystery word an ironic use of gæst ‘visitor, stranger’ or is it gæ¯st ‘demon, fiend’? The point is discussed in Hough and Corbett (2007: 120–124), who also note that Beowulf describes himself as a gæst, presumably ‘visitor, stranger’, in line 1800, while the monster Grendel is described as se grimma gæ¯st, presumably ‘demon, fiend’, in line 102. As modern readers, we can never be sure which meaning is intended in context; an Anglo-Saxon audience, listening to the poem rather than reading it, would have the difference in pronunciation to help them.
4.2 Metonymy and metaphor Two other kinds of semantic change which lead to polysemy are metonymy and metaphor, which have been a focus of study in semantics generally since the pioneering work of Lakoff and Johnson (1980). Following their lead, most work on this topic has been done within the framework of Cognitive Semantics, which draws on both linguistic and psychological theories of meaning. “Metonymy”, which many scholars consider to be the root of metaphor, usually occurs within semantically close areas of meaning when some aspect of an object or concept comes to refer to the whole, as when fa¯m ‘foam’ or wæ¯g ‘wave’ are used as synonyms for ‘sea’, or bord ‘plank, board’ is used to refer to a shield, ship or table, all of which are made of boards. In “metaphor”, words are transferred from one field of meaning to another, usually from concrete to abstract, as when ha¯t ‘hot’ from the field of physical temperature is transferred to the field of emotions, with meanings such as ‘fervent, excited’. From a diachronic point of view, one of the most interesting aspects of metaphor is its persistence through time. Sweetser (1990: 32–40) analyses metaphors of sense perception deriving from physical concepts, such as ‘grasping an idea’ or ‘seeing the truth’, which can be traced back to Indo-European, claiming that “[d]eep and pervasive metaphorical connections link our vocabulary of physical perception and our vocabulary of intellect and knowledge” (Sweetser 1990: 21). In Old English many words transfer from a meaning of physical vision to one of mental vision, including behealdan, bese¯on, lo¯cian, sce¯awian, all with a literal meaning of look at, gaze, and a metaphorical one of observe, regard, scrutinize. Kay (2000: 284) comments: “The Vision group of words incorporates an even more fundamental metaphor, that of holding/grasping or possession. Thus behealdan presumably follows an etymological path from holding in the hand to holding in the eye (that is seeing), to holding in the mind, that is understanding […] expressions for remembering include (ge)healdan, and habban/niman/lettan on gemynde”. Both Trim
20 Old English: Semantics and lexicon (2007) and Allan (2009) offer further insights into the evolution of metaphor. A good deal of work on the development of metaphor and metonymy within various theoretical frameworks has been done by G. A. Kleparski and his students at the University of Rzeszo´w, for example Kleparski (1990). Not all metaphors survive, however. In some cases, the metaphorical connexion remains even if it is differently lexicalized at various stages of a language, as when French fine replaces OE ðynne ‘thin’ in describing delicacy of perception. In other cases, the metaphorical connexion itself is lost, as happened to a group of metaphors for the body, mostly poetic compounds, where ba¯n ‘bone’ is followed by a word denoting some kind of container, as in ba¯ncofa ‘chest’, ba¯nfæt ‘vessel’, ba¯nhu¯s ‘house’, ba¯nsele ‘hall’. Containers and their properties, however, continue to supply metaphors, especially for the mind, as shown in an influential paper by Reddy on “The Conduit Metaphor” (Reddy 1979). Modern English examples include expressions like ‘the thought entered my head’, paralleled in Old English by uses of cuman/irnan on gemynde/on mo¯d ‘come to mind, occur to one’; hweorfan literally ‘turn’, metaphorically ‘turn the mind to’; bewindan literally ‘wind, wind round’, metaphorically ‘revolve in the mind’. Such examples show the underlying continuity of human conceptual processes even when, as in the case of hweorfan, the word itself has been lost.
5 The nature of the evidence Many of the problems encountered in studying the OE lexicon arise from the nature of the available data. Old English was spoken and written for over 600 years, with consequent diachronic, diatopic, and stylistic variation, but our evidence for such variation is patchy. Smith (1996: 17–19) notes how the survival of materials in the four generally recognized OE dialects, Northumbrian, Mercian, Kentish, and West Saxon, correlates with periods of historical importance for the areas concerned, and comments: “Apart from West Saxon, the dialect materials from Anglo-Saxon England are slight and fragmentary, and major parts of the country are almost entirely unrepresented (e.g. East Anglia)”. The majority of surviving texts, including the considerable body of poetry, are in West Saxon, which flourished along with the kingdom of Wessex in the 10th and 11th centuries. Late West Saxon is widely used as a model in grammars and dictionaries, and has been chosen as “the preferred spelling for headwords in the Dictionary of Old English (DOE)” (Healey 2006: 78). However, as Hogg (1992: 20) points out, while OE dialect features can be identified, “[…] there is almost complete social homogeneity between texts. Virtually every linguistic item we possess must have come from a very narrow social band indeed”, that is the small number of literate people. The extent of the problem of unrepresentativeness can be seen by a glance at the section below from TOE (Roberts and Kay 2000): 01.01.02.01.04.01 Marsh, bog, swamp: gebræ¯c, cwabbao, fenn, fengela¯dop, fenhleoþuop, fenhopop, fenland, fle¯othamq, fynig, gyr(u), gyrwefenno, hop, læc(e)q, mersc, merschopo, mersclando, mo¯r, mo¯rhopop, mos, piduq, polraq, sæ¯geq, slæd, snæpo, stro¯d, stro¯dettq, sucgaq, sumptq, wæsseq, wereþq […] Quicksand: cwecesondog, sandgeweorpg, sandridog (Roberts and Kay 2000: 7)
321
322
III Old English It will be observed that the majority of words are followed by one or more superscript flags which give a rough indication of the currency of the words (as opposed to particular meanings in the case of polysemous words). These are ‘o’ indicating infrequent use, ‘g’ for words occurring only in glossed texts or glossaries, ‘p’ for poetic register, and ‘q’ for doubtful forms. The flags are explained more fully in Roberts and Kay (2000: xxi– xxxi) where the authors state: “The flags point to aspects of word frequency that should always be held in mind, given that the extant corpus of Old English is small and probably skewed in its representation of Anglo-Saxon vocabulary”. Whereas the relatively small numbers of grammatical patterns in a language can be captured in a limited body of texts, parts of the larger and less stable corpus of lexical items may disappear wholly or partly from the record simply because the texts containing them are lost. In any lexical analysis, but especially in historical lexicology, frequency and context need to be taken into account. On the other hand, where evidence is scarce, any that survives must be of value. One area where we have a relatively large body of surviving texts is poetry (see Fulk, Chapter 25). Discussing traditional OE poetic diction, Godden (1992: 494) writes: “In both diction and syntax verse differs strikingly from contemporary prose and, one must assume, from contemporary speech”. As well as the compounds discussed in Section 3.2 above, poetic diction included simplex words not found in prose, such as Fre¯a and Metod as terms for ‘God’, and beorn and guma as terms for ‘man’. Such poetic words as naca ‘ship’, ga¯r ‘spear’, and wine ‘friend’ have prose equivalents in scip, spere, and fre¯ond. It is interesting, but perhaps not surprising, that the prosaic words, which are more likely to have been used in speech, are also more likely to survive into later stages of the language. It is also of interest that polysemous words could have both a poetic and a prosaic meaning. Thus Godden (1992: 498) notes: “[…] lind and helm are in general use in the senses ‘lime tree’ and ‘helmet’ but limited to poetry in the senses ‘shield’ and ‘protector’ ”.
6 Word and field studies Research on the OE lexicon takes many approaches, focussing on areas such as word structure (see Section 3), attempts to elucidate the meaning of individual words, and analyses of semantic fields and sub-fields. Much of this research is necessarily and desirably interdisciplinary in character, drawing on subjects such as archaeology, anthropology, and cultural history, as well as linguistics. A substantial body of lexicographical work at the interface of grammar and semantics has been carried out at universities in Spain using the Functional-Lexematic Model. Examples of this approach, and of others, can be found in Diaz Vera (2002). Earlier semantic field studies are listed and discussed in Strite (1989) and Kastovsky (1992: 400–407). There is space here to mention only a few of the studies which have been carried out since then; for additional information, the reader is referred to the bibliographies and reviews in that invaluable help to Anglo-Saxonists, the Old English Newsletter (see: http://www.oenewsletter.org/OEN/index.php). An example of a very focussed study is Schwyter (1996), which brings legal knowledge to bear on a primarily linguistic study of the lexical field of theft. The considerable body of work on kinship systems and terminologies carried out by anthropologists and ethnographers has
20 Old English: Semantics and lexicon informed work on the bifurcate OE kinship system, which largely disappeared after the Norman Conquest, for example in Fischer (2002). Cross-cultural approaches also provide many interesting insights in Anderson’s (2003) book on folk taxonomies for categories including color, seasons of the year, shapes, the five senses, psychology, and plant and animal life forms. Topics which have been particularly well-served in recent years include color, dress, food, and plants. Two notable interdisciplinary studies are Biggam (1997, 1998), which provide detailed analysis of terms for the concepts of ‘blue’ and ‘grey’ in Old English, drawing, for example, on botany and mineralogy to help identify the colors attached to plants and gem-stones respectively. She comments with feeling: “The basis of interdisciplinary semantics is the belief that the semantic study of a dead language needs all the help it can get” (Biggam 1997: 27). In another major study, Owen-Crocker (2004) uses evidence from a wide range of sources including archaeology, art, literature, and historical documents such as wills to build up a picture of Anglo-Saxon clothing, textiles and ornaments, thereby helping to elucidate the meanings of the words used to describe them. Various aspects of the production, processing and consumption of food are discussed in Hagen (1992, 1995) and Banham (2004). Plant-names, which are notorious for their uncertain meanings and diatopic variation, are the subject of two major projects. The Anglo-Saxon Plant-Name Survey (ASPNS) at the University of Glasgow aims to elucidate unknown names and review earlier studies; the proceedings of its first symposium appear in Biggam (2003). A joint project of the Universities of Graz and Munich has produced an online Dictionary of Old English Plant Names, building on the extensive previous work of Peter Bierbaumer, one of its editors (Bierbaumer et al. 2007–09). Given the increasing availability of online resources and electronic publication, it is to be hoped that many more such projects will be undertaken in future.
7 References Allan, Kathryn. 2009. A Diachronic Approach to Metaphor and Metonymy: How Target Concepts for Intelligence are Cognitively and Culturally Motivated. Oxford: Blackwell. Anderson, Earl R. 2003. Folk-taxonomies in early English. Madison, NJ: Fairleigh Dickinson University Press. Banham, Debby. 2004. Food and Drink in Anglo-Saxon England. Stroud: Tempus. Baugh, Albert C. and Thomas Cable. 1993. A History of the English Language. 4th edn. London: Routledge. Bierbaumer, Peter and Hans Sauer, with Helmut W. Klug and Ulrike Krischke (eds.). 2007– 09. Dictionary of Old English Plant Names. http://oldenglish-plantnames.uni-graz.at; last accessed 23/09/2009 Biggam, C. P. 1997. Blue in Old English: An Interdisciplinary Semantic Study. Amsterdam/Atlanta: Rodopi. Biggam, C. P. 1998. Grey in Old English. London: Runetree Press. Biggam, C. P. (ed.). 2003. From Earth to Art: The Many Aspects of the Plant-World in Anglo-Saxon England: Proceedings of the First ASPNS Symposium, University of Glasgow, 5–7 April 2000. Amsterdam/New York: Rodopi. Breeze, Andrew. 2002. Seven types of Celtic loanword. In: Markku Filppula, Juhani Klemola, and Heli Pitka¨nen (eds.), The Celtic Roots of English, 175–181. Joensuu: University of Joensuu.
323
324
III Old English Cameron, Angus, Ashley Crandall Amos, Antonette diPaolo Healey et al. (eds.). 1986–. Dictionary of Old English (DOE).University of Toronto. http://www.doe.utoronto.ca/about.html Cameron, Angus, Ashley Crandall Amos, Antonette diPaolo Healey et al. 2007. Dictionary of Old English: A to G. Toronto: Dictionary of Old English Project. http://www.doe.utoronto.ca/pub/ fasc-a-g.html; last accessed 21/08/2008 Clark Hall, J. R., with supplement by H. D Merritt. 1960. A Concise Anglo-Saxon Dictionary. 4th edn. Cambridge: Cambridge University Press. Diaz Vera, Javier E. (ed.). 2002. A Changing World of Words: Studies in English Historical Semantics and Lexis. Amsterdam/New York: Rodopi. Esposito, Anthony. 2002. Revising the coverage of Old English in the OED. http://www.oed.com/ newsletters/2002-06/old.html; last accessed 16/01/09 Fischer, Andreas. 2002. Notes on kinship terminology in the history of English. In: Katja Lenz and Ruth Mo¨hlig (eds.), Of Dyuersitie & Chaunge of Langage: Essays presented to Manfred Go¨rlach on the Occasion of his 65th Birthday, 115–128. Heidelberg: Winter. Godden, Malcolm R. 1992. Literary language. In: Hogg (ed.), 490–535. Hagen, Ann. 1992. A Handbook of Anglo-Saxon Food: Processing and Consumption. Pinner, Middlesex: Anglo-Saxon Books. Hagen, Ann. 1995. A Second Handbook of Anglo-Saxon Food: Production and Distribution. Hockwold cum Wilton: Anglo-Saxon Books. Healey, Antonette diPaolo. 2006. Straining words and striving voices: Polysemy and ambiguity and the importance of context in the disclosure of meaning. In: John Walmsley (ed.), Inside Old English: Essays in Honour of Bruce Mitchell, 74–90. Oxford: Blackwell. Hogg, Richard M. 1992. Introduction. In: Hogg (ed.), 1–25. Hogg, Richard M. (ed.). 1992. The Cambridge History of the English Language. Vol. I. The beginnings to 1066. Cambridge: Cambridge University Press. Hough, Carole and John Corbett. 2007. Beginning Old English. Basingstoke/New York: Palgrave Macmillan. Kastovsky, Dieter. 1992. Semantics and vocabulary. In: Hogg (ed.), 290–408. Kay, Christian. 2000. Metaphors we lived by: Pathways between Old and Modern English. In: Jane Roberts and Janet L. Nelson (eds.), Essays on Anglo-Saxon and Related Themes in Memory of Dr Lynne Grundy, 273–285. London: King’s College London Medieval Studies. Kay, Christian, Jane Roberts, Michael Samuels, and Irene´ Wotherspoon. 2009. The Historical Thesaurus of the Oxford English Dictionary. Oxford: Oxford University Press. Kleparski, G. A. 1990. Semantic Change in English: A Study of Evaluative Developments in the Domain of HUMANS. Lublin: University Press of the Catholic University of Lublin. Lakoff, George and Mark Johnson. 1980. Metaphors We Live By. Chicago: University of Chicago Press. Owen-Crocker, Gale R. 2004. Dress in Anglo-Saxon England. Woodbridge: Boydell. Reddy, Michael. 1979. The conduit metaphor. In: Andrew Ortony (ed.), Metaphor and Thought, 284–324. Cambridge: Cambridge University Press. Roberts, Jane and Christian Kay with Lynne Grundy. 2000. A Thesaurus of Old English. Amsterdam/Atlanta: Rodopi. Schwyter, J. R. 1996. Old English Legal Language: The Lexical Field of Theft. Odense: Odense University Press. Simpson, John A. and Edmund S. C. Weiner (eds.). 1989. Oxford English Dictionary. 2nd edn. Oxford: Oxford University Press. Simpson, John (ed.). 2000–. The Oxford English Dictionary. 3rd edn. online. Oxford University Press. www.oed.com/ Smith, Jeremy. 1996. An Historical Study of English: Function, Form and Change. London: Routledge. Strite, Vic. 1989. Old English Semantic Field Studies. New York: Peter Lang.
21 Old English: Pragmatics and discourse
325
Sweetser, Eve E. 1990. From Etymology to Pragmatics: Metaphorical and Cultural Aspects of Semantic Structure. Cambridge: Cambridge University Press. Trim, Richard. 2007. Metaphor Networks: Comparative Evolution of Figurative Language. Houndmills, Basingstoke: Palgrave Macmillan.
Christian Kay, Glasgow (UK)
21 Old English: Pragmatics and discourse 1 2 3 4 5
Old English discourse: data, texts and discourse communities Old English pragmatics as cross-cultural pragmatics Historical discourse analysis Summary References
Abstract This chapter suggests that OE pragmatics and discourse should be approached from a cross-linguistic and cross-cultural perspective, rather than seeing Old English as a prestage for the later periods of English. It is its cultural and linguistic “otherness” which makes Old English, in spite of the lack of good data, a particularly interesting area for pragmatic study. The different culture(s) of the Anglo-Saxon world required forms for the negotiation of meaning different from those we are familiar with today. Thus the conspicuous lack of structures attesting to politeness as face work, the existence of distinct speech events such as flyting or the prominent role of silence allow cross-cultural and cross-linguistic comparison which both corroborates and challenges issues like the uniformitarian principle. Similarly, the different typological character of the morpho-syntax of Old English allows a degree of word order flexibility that is exploited by discourse strategies.
1 Old English discourse: data, texts and discourse communities Pragmatics focuses on how meaning is negotiated, i.e., how speakers and hearers in certain contexts – to echo the title of Austin’s (1975 [1962]) groundbreaking lecture – “do things with words”. It thus concerns the analysis of mental processes in speakers and hearers, but also issues of linguistic and social interaction in specific socio-historical and cultural settings. While it is generally, also for speakers and hearers of Present-day languages, hard to isolate the crucial cognitive processes operating in the human minds, the study of OE pragmatics is complicated by at least two further factors: not only, as for all early periods of a language, the lack of good data (see Taavitsainen, Volume 2, Chapter 93), but also the length of the period and consequently and more importantly for the present subject, the changing linguistic and socio-historical conditions during and after the Anglo-Saxon period, which fundamentally affected the bases for and principles of social as well as linguistic interaction. Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 325–340
326
III Old English The OE period – traditionally considered to last from the middle of the 5th century to about 1100/1150 – is the longest among the conventional periods of English and covers more than 600 years (see von Mengden, Chapter 2), from the time when Germanic tribes, as pagan pirates and mercenaries, invaded Britain to the late Anglo-Saxon England of the 10th and 11th centuries, when the Anglo-Saxon society was one of the most sophisticated societies of the medieval West, renowned for its ecclesiastical, literary, and cultural achievements. Fortuitously, a wide variety of vernacular OE texts – many more than from any of the other early medieval Germanic societies – are extant from the Anglo-Saxon period. The online database Dictionary of Old English Web Corpus (Healey [ed.] 2005), which consists of at least one copy of every extant OE text, comprises about four million words of Old English, starting with Æthelberht’s vernacular law code from the late 6th century to a collection of diverse texts produced during and after the Benedictine Reform in the 10th and 11th centuries (see also the about 600 – a surprisingly large number – OE words designating textual categories and speech acts collected in Go¨rlach 2004: 91–97). Yet despite this exceptionally good preservation of data, it is still hard to study pragmatic and discourse patterns of the vernacular: Anglo-Saxon England may, at least for the majority of Anglo-Saxons, during its whole period be characterized as an oral rather than a literate society (see Schaefer, Volume 2, Chapter 81), but all of our extant texts are, of course, in the written medium, and all of them are strongly linked to the monastic settings in which most of the manuscripts were produced; accordingly, all of them are strongly influenced by a long literary tradition in Latin. When studying OE pragmatics and discourse, we thus have to be aware – much more than in later centuries – that we have access only to a very small proportion of the language actually used in Anglo-Saxon England. Furthermore, if we take a narrow approach and understand discourse as the spoken equivalent of a text, i.e., a stretch of conversation or dialogue, and discourse analysis as the examination of, for instance, patterns of turn-taking in a dialogue, its methods cannot be fruitfully applied to our OE material. For older stages of a language which are only extant in the written medium, it has been suggested that much can be deduced from socalled “speech-based genres” such as court records, drama or from more colloquial written genres such as personal letters or diaries (see Biber and Finegan 1992). Yet virtually no such speech-based genres are extant from the OE period, except for the didactic dialogues found in the OE version of Ælfric’s Colloquy and the indirect instances of direct speech in OE fictional texts attested in poetry, homilies, or prose narratives such as the OE Apollonius. Most of the homilies and narratives, however, are translations from Latin, so that we cannot be sure whether the speech conventions recorded there echo actual OE speech interaction or whether they were typical of the Latin discourse tradition or a hybrid Anglo-Saxon/Latin tradition. In his study of interjections in the OE part of the Helsinki Corpus (Rissanen et al. 1991), for instance, Hiltunen (2006: 92) finds that OE eala, the most frequent interjection in the corpus (66 instances), is attested in a broad variety of texts as an attention-getter or emphasizer, but only in texts translated or adapted from Latin Christian models. It is thus an OE interjection with decidedly literary, especially Christian, associations (for a similar text-specific distribution of OE insulting epithets, see Chapman 2008 and Section 3.2). Even in the more monolingual context of heroic poetry (which survives in revised Christianized form only) and other instances of oral formulaic poetry (see Schaefer, Volume 2, Chapter 81), such as examples of AngloSaxon verbal duelling, we cannot be sure whether we are dealing with actual language
21 Old English: Pragmatics and discourse interaction or merely literary topoi typical of the literary traditions in the vernacular or in Latin (see the discussion of “flyting” in Knappe 2008 and Section 2.2 below). If we take a broader view of the concept of discourse and capture it as a domain of communication denoting the totality of linguistic practices that pertain to a particular field of knowledge (such as, for instance, the discourses of the courtroom, of law, of news, of science) and the dissemination of information within a certain group of speakers, i.e., a specific discourse community (communities with a common set of assumptions and a shared discourse), the only promising field for Old English is the “discourse of religion” (see also Kohnen 2007a), which is widely attested in translations of the psalms and the gospels as well as translations of works by the Fathers (Gregory the Great, Augustine, Bede) and in text types such as prayers, homilies, monastic rules (St. Benedict, Chrodegang) or penitentials. It is exactly the emergence of new text types and forms of discourse which characterizes the Middle English period (see Traugott, Chapter 30).
2 Old English pragmatics as cross-cultural pragmatics This survey of existing texts and their contexts shows that a comprehensive study of OE pragmatics and discourse would have to cover as diverse concepts as the discourse traditions of the Germanic heroic age and those of a recently Christianized society, and also the scholarly activities in the vein of the Benedictine reform (the last two strongly influenced by Latin literacy and its discourse traditions). Anglo-Saxon culture – or rather, Anglo-Saxon cultures – thus was very much different from later cultures: it not only saw the transition from an oral to a literate society, but also – an aspect relevant for the major principles of social interaction central to pragmatic analysis – the transition from a heroic to a Christian society. The Anglo-Saxon period thus has very different medial and cultural backgrounds compared to later periods of English. Yet the period is sometimes in danger of losing what has come to be called its essential “alterity”, its otherness compared to our habits of mind, modes of expression and principles of social organization (see Jauss 1979; Lerer 1991: 7-8). Very little of this otherness, for example, seems to have been acknowledged in studies of OE pragmatics, most probably because many of the studies are based on the OE corpus material selected for the Helsinki Corpus (Rissanen et al. 1991), which aims at the comparability of genres and text types during the history of English and not their divergence.
2.1 Politeness Some issues of divergence or otherness are enunciated in Kohnen’s more recent discussions of the question whether there was anything like facework in terms of the politeness theory by Brown and Levinson (1987) in Anglo-Saxon England and if so, whether these norms were more oriented towards positive or negative politeness (Kohnen 2008a, 2008b). In his studies on different manifestations of OE directives (Kohnen 2000, 2007b, 2008a, 2008c; see also Section 3.1.2), Kohnen finds no instances of negative politeness, i.e. the wish that one’s actions go unimpeded by the others, in texts set in Germanic or secular contexts. Instead, direct performatives Ic bidde eow þæt … ‘I ask you to …’ or constructions with þu scealt ‘thou shalt’ are preferred. In texts set
327
328
III Old English within a Christian context, many strategies implying a basic kind of solidarity are attested, such as constructions with hortative uton ‘let us’, stressing the necessity of the required action from both addresser and addressee. Kohnen argues that the solidarity expressed by common ground strategies reflects the Christian and monastic models of ‘humilitas’ and ‘oboedientia’ which, in his opinion, cannot necessarily be taken as strategies of face work (Kohnen 2008b: 143). These findings suggest that linguistic politeness in the sense of face work may not have been important in Anglo-Saxon communication, at least with regard to negative politeness. A similar distribution across text categories is suggested by Kohnen’s study of OE terms of address, such as leof ‘dear one, friend’, broþor ‘brother’, or hlaford ‘master, ruler; lord’. Leof, the most general courteous address in Old English communicating sympathy, affection, and friendship, is neither typically formal nor typically authoritative, but applicable in a wide range of relationships and settings (so that the usual Modern English translations Sir or My/Dear Lord, which imply a certain authoritative hierarchy, are inaccurate). Broþor designates a friendly and affectionate relationship and seems to combine the intimate, mostly affectionate bond associated with blood relationship and the basic solidarity among humans being requested by Christian morals (Kohnen 2008b: 145– 152). In secular texts, the use of hlaford reflects a static hierarchical society, a fixed rank in a hierarchical society characterized by mutual obligation and kin loyalty; in religious contexts, it is used to address God or Christ (translating Lt. dominus, i.e. ‘the Lord’). Kohnen summarizes that the prevalent picture of a warlike society of (secular) Anglo-Saxon England may have followed different underlying assumptions and customs, suggesting that face-threatening acts were not felt as a menace but rather as an accomplishment and that face-enhancing acts, like self-praise and boasting (see Section 2.2 on flyting), were not considered to be embarrassing. Politeness as face work may thus not have played a major role in Anglo-Saxon society. This highlights the intrinsically culture-specific nature of phenomena like politeness and suggests in accordance with other cross-cultural studies that the universal validity or significance of politeness theory – as devised by Brown and Levinson (1987) – is a gross mistake. Negative politeness in particular is fundamentally culturespecific, reflecting the typical patterns of today’s Western, or even more particular, Anglo-American, politeness culture (see also Jucker and Taavitsainen 2008: 7–9). The study of Anglo-Saxon pragmatics thus does not only affect our understanding of the historicity of verbal interaction but also challenges issues of universality.
2.2 Flyting Similar factors of “otherness” have also been studied – in literary as well as pragmatic investigations – in the analysis of Anglo-Saxon “flyting” (cf. OE flitan ‘strive, quarrel, dispute’), the defiant, proud provocation in verbal duelling in heroic poetry (see Arnovick 1999: 15–40; Jucker and Taavitsainen 2000; Knappe 2008). In particular Byrhtnoth’s flyting in the Battle of Maldon (lines 25–61) or the so-called Unferth episode in Beowulf (lines 499–610) exemplify the Germanic genre with its highly stylized and conventional rhetoric (following the standard sequence “claim – defence – counterclaim”) and also the subtlety of the combination of insult and boast. In flyting, power and status are negotiated on a verbal battlefield; typical topics of the insults are crimes of kinship (cowardice, failure of honor, irresponsible behavior). In the Unferth episode,
21 Old English: Pragmatics and discourse for instance, Beowulf is invited to tell of his famous victories, but first Unferth addresses Beowulf with an insulting speech, accusing Beowulf of having risked his life for a foolish contest with Breca and for having lost the contest (lines 506–511). In the passage given, (1), Unferth provocatively doubts that Beowulf is going to be successful in his encounter with Grendel (lines 512–515) and Beowulf counters in the appropriate style of flyting by accusing Unferth of being drunk (see Arnovick 1999: 608–609 and Jucker and Taavitsainen 2000: 77–78). (1)
[Unferth] “Đonne wene ic to þe wyrsan geþingea ðeah þu heaðoræsa gehwær þohte grimre guðe, gif þu Grendles dearst nihtlonge fyrst nean bidan.” Beowulf maþelode, bearn Ecgþeowes: “Hwæt, þu worn fela, wine min Unferð beore druncen ymb Brecan spræce sægdest from his siðe!” (Beowulf 525–532a) ‘Unferth: “Therefore I anticipate worse outcome for you – though you may always have proved competent in the onslaughts of battle and fierce fighting – if you dare to await Grendel at close quarters for the duration of a night”. Beowulf, son of Ecgtheow, spoke out: “Well now, Unferth, my friend, you have a lot to say about Breca and to tell about this enterprise for one who is drunk with beer!” ’
Ritual insults like these continue beyond the period of heroic poetry in the literary challenges between later medieval knights or its revival as a Scots literary genre in the Renaissance. In cross-cultural approaches, its characteristics have recently also been compared to the – also predominantly oral – ritual insults in the sounding or playing the dozens by African-American adolescents. Both flyting and sounding can be described as rule-governed and therefore ritual, but Anglo-Saxon flyting, arguably (see the discussion in Knappe 2008), lacks the ludic character of the sounding of urban black adolescents in the English-speaking world (for this distinction, see Jucker and Taavitsainen 2000: 77).
2.3 Frame analysis: Old English charms Studies like these call attention to Bax’s (2001) suggestion that comparative frame analysis as conceived by Goffman (1974) may be an effective device for an analysis of the “otherness” of distinct medieval speech events such as ritual challenging. Yet until now very little use has been made of the historical dimensions of frame analysis, such as the recognition of “scripts” or “frames”, i.e., pre-existing knowledge structures for interpreting event sequences with a fixed static pattern (see Yule 1996: 85–89). Related ideas and approaches have, however, been applied in the studies on Anglo-Saxon “flyting” summarized in Section 2.2 and particularly in Arnovick’s (2006) investigation of verbal performatives in a corpus of 463 OE and Anglo-Latin charm incantations from the Lacnunga and the Leechbook. In an interdisciplinary approach inspired by both historical-pragmatic and oral theory, charm incantations are seen as “speech events”, i.e. culturally recognized social activities within a very particular social and linguistic context in which language plays a specific role.
329
330
III Old English The speech event charm incantation is characterized by a ritual communication with spirits. Arnovick’s analysis concentrates on elements with a clear pragmatic function, most of which cannot be studied by corpus methods because they comprise words with non-propositional meaning (“Speaking Gibberish”; Chapter 2), foreign language elements (“Praying the Pater Noster”; Chapter 3), or use no words at all (“Keeping Silence”; Chapter 5). In her close readings of these charms, Arnovick shows that gibberish utterances, which seem to address the spirits in their own tongue, are used as performative relics which perform the illocutionary work of the charm. Gibberish is thus essentially performative because it is the medium of word magic. Similar functions are established for the Latin Pater Noster and for silence: If the Pater Noster follows a gibberish utterance, it serves to sanctify its command; if the Pater Noster appears as the single incantation in the charm, it shoulders the whole illocutionary burden of this charm. Silence – as a metacommunicative marker of the incantation, introducing or framing the magical utterance – does not only signal respect to the deity addressed, but it also indicates respect to the audience of the ritual. Arnovick’s analyses, which are also based on a detailed analysis of contextual settings as testified by contemporary texts such as monastic Rules, are a pertinent example of how such a close analysis may depict past and no longer existing models of verbal interaction, which thus expose the “alterity” of Anglo-Saxon social and linguistic interaction.
3 Historical discourse analysis In addition to these pragmaphilological and interdisciplinary approaches, we find various investigations in the scope of historical discourse analysis (for the term, see Brinton 2001: 139), both function-to-form and form-to-function mappings. Function-to-form mapping, i.e. the identification of OE forms which realize particular discourse functions, is relevant both for larger frames or scripts such as flyting (see Section 2.2 above) or the investigation into certain speech acts (see Section 3.1 below). Form-tofunction mapping, i.e. the explication of pragmatic and discourse functions of a particular OE item, is central for the analysis of, for example, speech act verbs such as biddan and beodan, which are used as performative directives (also see Section 3.1.2) or for investigations into the functions of polyfunctional items such as þa, soþlice, or hwæt, which have been analyzed as discourse markers in Old English (see Section 3.3). If we survey these recent contributions to OE discourse analysis, however, we cannot fail to notice a conspicuous ad hoc character of most of the studies and their rather large conceptual variety and methodological pluriformity. Again, most of these studies prefer analyzing features and patterns attested for Present-day English, sometimes neglecting the fact that Old English differentiated – as a still much more inflecting language – more nominal and verbal categories (grammatical gender, case; mood), which allow the pragmatic exploitation of structures which are no longer possible in Present-day English. For directives, Old English could employ not only imperative, but also subjunctive/optative inflections (Traugott 1991: 398), instead of or in addition to the employment of various speech act verbs comparable to, for example, PDE order (OE beodan) or ask (OE biddan). Similarly, tense-aspect morphology and, in particular, the employment of specific different patterns of word order to signal focus or topic relations (which are no longer applicable in the fixed system SVO in Present-day English), play an important role in foregrounding and narrative segmentation in Old English, instead of or in addition to discourse markers such as þa, hwæt, or soþlice (see Section 3.4).
21 Old English: Pragmatics and discourse
3.1 Function-to-form mapping: speech acts 3.1.1 Methodological issues In the beginnings of historical pragmatics following Jucker’s (1995) landmark volume, the study of speech act verbs and speech acts was considered to be particularly promising (see the inaugural issue of the Journal of Historical Pragmatics 2000) and so a number of studies on English diachronic pragmatics have dealt with speech acts (for the methodologically most consistent approaches, see Jucker and Taavitsainen 2008 [eds.]). Apart from Kohnen’s intense synchronic investigations into OE directives (see Section 3.1.2), most of these studies aim at comparing certain speech acts through the history of English; see Traugott (1991) on the history of English speech act verbs, Arnovick (1994, 1999: 57–94) on promising and curses, Jucker and Taavitsainen (2000) on insults, and Grzega (2008) on greetings. Accordingly, their primary focus is on regular patterns in language change, such as regularities in the semantic shift from non-epistemic to epistemic in the development of speech act verbs (Traugott 1991) or an increasing subjectification in promising or cursing (Arnovick 1994, 1999). Much of this research is not only inhibited by the lack of comparative data, but also – as Bertuccelli Papi (2000) has pointed out – by the theoretical divergence of the Austinian, Gricean, and Searlian traditions, which emphasize the detailed investigation of the socio-historical context, on the one hand, or, on the other, an essentially cognitive notion of context, i.e., speaker intentions, felicity conditions, and mental attitudes (which are particularly hard to reveal for OE speakers). Many of the studies on English historical pragmatics employ a mixed approach, which is, however, basically structured along the lines of the five types of speech acts distinguished by Searle (1969; representatives, directives, commissives, expressives, and declaratives). Because of the conspicuous synchronic and diachronic variation in speech acts, Jucker and Taavitsainen (2000, 2008) have suggested a prototype approach, viewing speech acts within the “multidimensional pragmatic space” they share with neighbouring speech acts whose coordinates are context-specific, culture-specific and time-specific. The pragmatic space of “antagonistic verbal behaviour” (Jucker and Taavitsainen 2008), for example, would thus include insults and threats as found in medieval flyting, Shakespearian name-calling, or present-day sounding or flaming, since basic patterns of the speech act “verbal insult” are repeated there in slightly modified forms (see Section 2.2 above). Studies following this approach have thus confirmed the importance of long-term investigations in historical pragmatics, since they both corroborate and challenge issues like the uniformitarian principle (for the challenges, see also Section 2.1 on politeness, above). Again, however, it has to be acknowledged that it is in particular divergence which is crucial for the study of OE speech acts. In her pioneering account outlining the chances and challenges of historical pragmatics, Schlieben-Lange (1976) warned that we have to be very careful when transferring our understanding of today’s speech acts and their principles to older stages of a language. Not only were there different, or at least much more important and more highly institutionalized, speech acts, such as BANISHING, OUTLAWING, SCORNING, OFFERING ONE’S SERVICE, but some speech acts and events which seem equivalent to present-day ones may have changed in their pragmatic function. In his study of the Old High German performative formula in the speech event of baptism, Wagner (1994), for example, accordingly finds that the pragmatic functions of certain performatives have changed tremendously: while in the medieval theocentric world view of
331
332
III Old English the early Middle Ages, the utterance of performative formulas employing the verbs gelouban ‘believe’ and forsahhan ‘forsake’ was an act of abjuring the devils and subjecting oneself to the Christian god, i.e. a declarative speech act in Searlian terms (speaker causes the situation X, direction of fit: words change the world), it is – for many speakers – now rather a solemn profession of faith or a solemn promise and therefore a commissive speech act (the speaker intends X, direction of fit: make the world fit words).
3.1.2 Directives in Old English In general, Schlieben-Lange (1976) suggested that indirect speech acts, ambiguous in their illocution or perlocution, were rare, reflecting the different social contexts in a strongly hierarchical society. For Old English, these suggestions were corroborated for directives and their development, the as yet only systematically studied OE speech act (see Kohnen 2000, 2008a, 2008c). In an early study, Kohnen (2000) found that performative directives, i.e., instances in which the speaker explicitly refers to the act of requesting or commanding, were more common in Old English than in Present-day English, where directives are often realized in indirect or hedged form (e.g. Could you give me a hand? or Will you do me a favour?). The following examples, (2a, b) from Ælfric’s letter to Wulfsige and the OE laws of King Canute illustrate the “typical explicit performative pattern”, i.e. constructions with verbs in 1P SG active, (preferentially) an object referring to the addressee, and a subordinate clause introduced by þæt naming the requested action: þæt
(2) a. And we beodað and
man Cristene men
we command-1P.PL.ACT that [SUB. CONJ.] one
to lytlum
huru
to deaþe
for ealles
Christian men-ACC.PL for all–GEN.SG ne
forræde
to little (things)–DAT.PL certainly to death-DAT.SG not sentence–3P.SG.ACT.SUBJ ‘And we command that Christian men be not sentenced to death for the slightest reason’ (c.1020 LawIICn, 2.1; Kohnen 2000: 304) b. ic bidde I
eow
þæt
ge
gymon
ask-1P.SG.ACT you-2P.PL.ACC that [SUB. CONJ.] you-2P.PL.NOM take care–2P.PL.ACT.SUBJ
eowra
sylfra
swa eowere
your-2P.PL.GEN selves-2P.PL.GEN as
bec
eow
wissiað
your-2P.PL.GEN books–NOM.PL you-2P.PL.ACC instruct-3P.PL.ACT
‘I ask you to take care of yourselves in such a way as your books instruct you’ (c.1000 ÆLet 1 [Wulfsige Xa] 117; Kohnen 2000: 304)
In Old English, the performative function was found to be restricted to only five out of the altogether 40 relevant speech act verbs, namely biddan ‘ask’, læran ‘teach’, halsian ‘implore’, bebeodan ‘bid’, and beodan ‘command’. This means that only those directive acts were prominent which imply an unambiguous and asymmetric relationship between addresser and addressee: Either the addresser holds a superior rank (as in the case of socalled beodan verbs such as beodan, bebeodan, hatan ‘command’, læran ‘instruct’, mynegian ‘exhort’, manian ‘exhort’) or the addresser is not in a superior position (as in the case of so-called biddan verbs such as biddan, gebiddan ‘ask’, halsian). Furthermore, the conspicuous lack of verbs denoting suggestion and advice such as PDE suggest or recommend shows that indirect directives and, consequently, the tendency to avoid face-threatening acts seem to have developed relatively late in the history of English, in Early Modern English (see Kohnen 2002, 2007b; see also Section 2.1, above).
21 Old English: Pragmatics and discourse In his following studies, charting the complete inventory of the various manifestations of directives, Kohnen finds that the manifestations of directives fall into four classes: performatives, imperatives, modal expressions, and indirect manifestations (see Kohnen 2007b and 2008c). Over the history of English, performatives, modals, and in particular imperatives, with 2P imperatives as their unmarked manifestation, have been most frequent (Kohnen 2008c: 309). The most problematic class of directives for all periods of English are indirect directives, in which an utterance contains neither imperatives, nor the relevant modals, nor performatives; in his OE data, Kohnen could not find any instances of such indirect directives (Kohnen 2008c: 301): both their frequency and their variability increase only over the centuries. All in all, the findings thus again suggest that negative politeness did not play a major role in Anglo-Saxon communication (see also Section 2.1).
3.1.3 Other speech acts In her study on the history of promises in English, Arnovick (1994, 1999: 57–71) similarly proposes a “straightforward nature of the promise” in Old English: promises in Old English are described as direct, sentence-length utterances; many of the promissory statements rely upon sculan or willan, such as in (3): (3)
Nu ic, Beowulf, þec, secg betsta, me for sunu wylle freogan on ferhþe. (Beowulf 946b–948a) ‘Now, Beowulf, best of men, in my heart I will love you as a son’
Arnovick further argues that the development of will and shall from deontic modals to epistemic tense markers, marking predominantly futurity and no longer obligation, results in an expansion of the discourse needed to convey the illocutionary force of promising in the course of the history of English. In the development of the manifestations of SWEARING/CURSING, Arnovick also finds a movement toward greater subjectivity from Old English, when curses were so standard as to appear formulaic and had an exclusively deontic, religious meaning (Arnovick 1999: 73–94 “Subjectification in the Common Curse”). Analogous stylized forms have also been found for the related speech act INSULT in Old English (cf. Jucker and Taavitsainen 2000 and the account of OE flyting, also Section 2.2 above).
3.2 Form-to-function mappings 3.2.1 Insulting epithets With respect to form-to-function mapping, i.e., the explication of pragmatic and discourse functions of particular OE items, a similar importance of stylized forms is seen in Chapman’s (2008) study of insulting epithets in Old English, which highlights the highly conventional character of insulting epithets. OE speakers/writers use common, well-worn words, which are frequently repeated in the epithets (earm and earmig ‘poor’ are most versatile; there are only four hapax, such as wambscyldig ‘belly-guilty; sinful’) and they fall into a fairly well defined set of semantic categories, typical of insults in other
333
334
III Old English languages, such as “low social standing” (e.g. earm/earmig ‘poor; miserable, wretched’, ungesælig ‘unfortunate; miserable, wretched’), “intellectual, mental deviations” (e.g. dysig ‘foolish’, stunt ‘foolish’), “individual deviations in character” (e.g. lyþre ‘wicked, evil’, wælhreow ‘cruel, barbarous’), or “individual transgressions of societal norms”, in Old English overwhelmingly characterized as sins (e.g. druncene ‘drunk’, leas ‘vain, false; lying’, wlanc ‘proud’, licettere ‘hypocrite’, wedloga ‘oath-breaker’). In sum, Chapman finds a preponderance of terms naming sins or other moral shortcomings, such as wedloga ‘oath-breaker’ or licettere ‘hypocrite’. Notably missing from the OE data are sexual and scatological epithets, which appear to have been common in Germanic traditions, but which might not be attested because of the predominance of religious genres in the surviving OE texts (Chapman 2008: 4). The most creative expressions are found in addresses to devils and in an exceptional passage in an OE debate between the body and the soul from the Vercelli Homilies, where the soul accuses the body of getting them both damned: (4)
La, ðu eorðan lamb & dust & wyrma gifel, & þu wambscyldiga fætels & gealstor & fulnes & hræw, hwig forgeate ðu me & þa toweardan tide? (HomU 9 [ScraggVerc 4] 207) ‘Hey, you mud of the earth and dust and food for worms, and you belly-guilty bag and pestilence and foulness and corpse, why did you forget me and the future?’ (translation Chapman 2008: 2)
While this passage may indeed allow a better idea of spoken OE interaction, the common conventionality of insults is not specific to Old English, but has already been noted for many other languages. Indeed, this speech act actually invites conventionality, because an inherent characteristic of an insult is that the insulting label must be recognized as such by the target and other listeners.
3.2.2 Interjections Single OE elements with pragmatic function which have received some attention in the literature are interjections and, in particular, discourse particles. Some approaches propose a cline between interjections and discourse markers (see the review of scholarship in Hiltunen 2006: 93–94) and there are a number of polyfunctional items such as OE hwæt ‘what’, which may function as interjections or discourse particles. The main difference between interjections and discourse markers, however, is that while the latter work on the textual and interpersonal level only, interjections predominantly function as full speech acts, i.e., as equivalents to a full sentence (PDE Wow! ‘I am surprised’, PDE sh ‘I want silence here’). The studies on interjections have until now mainly concentrated on charting the inventory of interjections in Old English by means of a corpus analysis (Hiltunen 2006) and investigations into the use and descriptions of OE interjections (and their relation to Latin ones) in OE texts and also metalinguistic sources such as Ælfric’s Grammar (Hiltunen 2006; Sauer 2007, 2008). An investigation of Ælfric’s Grammar shows that Ælfric was aware of the role and significance of interjections in English, since he discusses them in some detail. A more functional approach to interjections is likely to yield not only further insights into interjections but also, since they signal full speech acts, into speech acts and their analysis.
21 Old English: Pragmatics and discourse
3.2.3 Discourse markers Starting with Enkvist’s analyses of OE þa ‘then; when’ (Enkvist 1972, 1986) and, in particular, Brinton’s pioneering monograph on “mystery features” of Old and Middle English (Brinton 1996), most attention in OE pragmatic research has until recently been paid to what have been termed “discourse markers” (Schiffrin 1987) or “pragmatic markers” (Brinton 1996). Due to the difficulties of sentence and discourse segmentation arising from the lack of punctuation in OE manuscripts and also due to the oral or oral-literate character of OE texts in general, the discourse-structuring function of adverbs and phrases which segment OE texts was registered in philological approaches to Anglo-Saxon literature before the beginning of historical pragmatics as a discipline proper. OE adverbs such as her ‘here’ and nu ‘now’, for example, have been described in their text deictic, i.e., textual, functions (Clemoes 1985; Fries 1993; see also Lenker 2000). Much valuable material can also be found in glossaries and introductions to editions of OE texts or monographs discussing the language of individual Anglo-Saxon authors (for the Alfredian works, see, e.g., Wu¨lfing 1894–1901), material which tends to be neglected in studies based on language corpora such as the Helsinki Corpus (Rissanen et al. 1991). The OE items best studied for their various functions are þa ‘then; when’ and hwæt ‘what’, not only in linguistic but also in literary and philological studies hwæt in particular because of its prominent appearance at the beginning of several well-known OE poems such as Beowulf, Andreas, The Dream of the Rood, Exodus, Fates of the Apostles, Judgement Day II, Solomon and Saturn, and Vainglory (see Brinton 1996: Chapter 7 and Stanley 2000). Yet, with the advent of text linguistics and pragmatics, the investigations into these items have become much more methodological. Brinton’s pragmatic analysis of the uses of OE hwæt, for example, shows that it – in addition to its employment as an interrogative and complementizer – serves as an attention getter and as a marker of shared knowledge (Brinton 1996: 187–189). OE þa may be employed as a temporal adverb ‘then’ or as a temporal conjunction ‘when; then … when’. Within a discourse perspective, þa has been seen to function as a discourse marker denoting foregrounding action, narrative segmentation, or shifts on the discourse level (see, for example, Enkvist 1972, 1986; Enkvist and Wa˚rvik 1987; Kim 1992; Brinton 2006). In most of the studies, word order patterns (mainly verb-second, i.e., þa V…) are considered to analyze and highlight the various text-structuring function of þa (Wa˚rvik 1995, 2011; see also below, Section 3.4). Similar functions have been found for two collocations comprising þa: hwæt þa ‘what then’ moves the narrative forward, expressing the fact that the event that follows can be inferred from the previous event (Brinton 1996: 193–199). The phrases þa gelamp / gewearþ / wæs (hit) (þæt) ‘then it happened that’ – termed gelamp-constructions by Brinton (1996: Chapter 5) – serve as episode boundary markers, expressing the “subsidiary foreground”, the instigating event of an episode. Comparable to OE þa and OE þa gelamp hit, OE soþlice and witodlice, lit. ‘truly, verily’, may not only be employed as manner adverbs with a scope within the predicate or as truth-intensifying emphasizers, but, more often, as pragmatic markers functioning as highlighters or – denoting episode boundaries – as markers of discourse discontinuity in OE prose (Lenker 2000).
335
336
III Old English
3.2.4 Word order and information structure Recent work has highlighted the relevance of information structure to the choice of word order options in Old English. Again, it is the difference of Old English vs. Middle (and Present-day) English structures which has to be stressed: much more than in later periods of English, morphological features (tense-aspect inflection) and distinct word order patterns were central for text-structuring. Tense-aspect morphology, which serves the function of placing events in time (with aspect dependent on the speakers’ perspective), played an important role in foregrounding and narrative segmentation: In his analysis of fore- and backgrounding in the Anglo-Saxon Chronicle, Hopper suggests that the foreground in OE narratives, in accordance with general principles of grounding, is indicated by the perfective aspect (single dynamic, punctual, or telic events), whereas the background (durative/iterative/habitual/atelic processes) is indicated by the imperfective aspect. Foregrounding may also be indicated through VS and OV structures and backgrounding by SV structures (Hopper 1979: 220–226). The idea that discourse relations are signalled by specific word order patterns is thus not new, but research into the relation of information structure and word order has seen an upsurge of interest in recent years. So Los (2000) finds that different word orders with onginnan/beginnan ‘begin’ in the works of Ælfric yield identifiable discourse effects: on-/beginnan in V1 position invariably indicate discourse discontinuity, i.e. mark episodes in which the narrative takes a dramatic turn, whereas main clauses introduced by þa + finite on-/beginnan take a bare infinitive and mark discourse continuity. In her examination of word order patterns in Old English non-coordinate and coordinate clauses, i.e., clauses introduced by coordinating conjunctions, Bech (2008) finds that verb-final main clauses (SXV) are likely to signal coordinating discourse relations, for example Narration or Continuation. In a number of recent articles, van Kemenade and colleagues highlight the fact that Old English possesses a number of morpho-syntactic properties which allow a degree of word order flexibility that is exploited by discourse strategies and, in particular, suggest that particular adverbs (mainly þa and þonne ‘then’) functioned as “discourse partitioners” in Old English (van Kemenade and Los 2006; van Kemenade et al. 2008). Their examination of word order patterns with OE þa and þonne leads them to claim that these adverbs (or rather, particles) should be seen as “focus particles” with a fixed position in the clause structure, with a topic area to the left of the particle and a focus area on the right (for an alternative view on the discourse functions of adverbs in “post-first-position”, i.e. the position after the first sentence constituent, see Lenker 2010: 67–72). In sum, these studies suggest that Old English is tailored to allow a certain amount of discourse flexibility: the syntactic and – undisputed – discourse properties of þa/þonne show that the syntactic organization of the clause in Old English is closely interwoven with discourse organization, while the transition to Middle English is one that results in a more strictly syntactic organization of the clause (see also Traugott, Chapter 30).
4 Summary The studies of OE pragmatics and discourse reviewed here, though as yet very diverse in their approaches and methodologies, have shown that it is indeed possible to obtain an understanding of how meaning was negotiated in Anglo-Saxon times, in spite of the
21 Old English: Pragmatics and discourse lack of good, especially spoken or speech-based, data. It is in particular the “alterity” of Anglo-Saxon England culture(s), i.e. the very different socio-historical conditions, and the typological difference of the still largely inflecting structure of Old English as compared to Present-day English, which make Old English a very interesting field for the study of pragmatics. It might thus be advisable to approach OE pragmatics more like cross-cultural, contrastive studies, rather than viewing it – as many studies so far seem to have done – as a pre-stage for Middle English, Early Modern English, and Modern English. Contrastive cross-cultural approaches to speech acts (especially by Blum-Kulka and associates; see Blum-Kulka et al. 1989) should thus be methodologically significant for the study of OE pragmatics, since they compare the realization of a particular speech act in different cultural and linguistic contexts. Although many aspects of the history of larger speech events and speech acts may prove to be cultural ones in the final analysis, it is thanks to linguistic, i.e., pragmatic, studies of Old English that many complex stories of linguistic and cultural interaction in Anglo-Saxon England have already been exposed.
5 References Arnovick, Leslie K. 1994. The expanding discourse of promises in Present-day English: A case study in historical pragmatics. Folia Linguistica Historica 15: 175–191. Arnovick, Leslie K. 1999. Diachronic Pragmatics. Seven Case Studies in English Illocutionary Development. Amsterdam/Philadelphia: John Benjamins. Arnovick, Leslie K. 2006. Written Reliquaries: The Resonance of Orality in Medieval English Texts. Amsterdam/Philadelphia: John Benjamins. Austin, J. L. 1975 [1962]. How to do things with words. In: J. O. Urmson and Marina Sbisa` (eds.), The William James Lectures Delivered at Harvard University in 1955. 2nd edn. Oxford: Oxford University Press. Bax, Marcel. 2001. Historical frame analysis: Hoaxing and make-believe in a seventeenth-century Dutch play. Journal of Historical Pragmatics 2: 33–67. Bertuccelli Papi, Marcella. 2000. Is a diachronic speech act theory possible? Journal of Historical Pragmatics 1: 57–66. Bech, Kristin. 2008. Verb types and word order in Old and Middle English non-coordinate and coordinate clauses. In: Gotti, Dossena, and Dury (eds.), 49–67. Biber, Douglas and Edward Finegan. 1992. The linguistic evolution of five written and speech based English genres from the 17th to the 20th centuries. In: Matti Rissanen, Ossi Ihalainen, Terttu Nevalainen, and Irma Taavitsainen (eds.), History of Englishes: New Methods and Interpretations in Historical Linguistics, 688–704. Berlin/New York: Mouton de Gruyter. Blum-Kulka, Shoshana, Juliane House, and Gabriele Kaspar (eds.). 1989. Cross-Cultural Pragmatics: Requests and Apologies. Norwood, NJ: Apex. Brinton, Laurel J. 1996. Pragmatic Markers in English: Grammaticalization and Discourse Functions. Berlin/New York: Mouton de Gruyter. Brinton, Laurel J. 2001. Historical discourse analysis. In: Deborah Schiffrin, Deborah Tannen, and Heidi E. Hamilton (eds.), The Handbook of Discourse Analysis, 138–160. Oxford: Blackwell. Brinton, Laurel J. 2006. Pathways in the development of pragmatic markers in English. In: van Kemenade and Los (eds.), 307–334. Brown, Penelope and Stephen C. Levinson. 1987. Politeness: Some Universals of Language Use. Cambridge: Cambridge University Press. Chapman, Don. 2008. “You belly-guilty bag”. Insulting epithets in Old English. Journal of Historical Pragmatics 9: 1–19. Clemoes, Peter. 1985. Language in context: Her in the 890 Anglo-Saxon Chronicle. Leeds Studies in English 16: 27–36.
337
338
III Old English Enkvist, Nils Erik. 1972. Old English adverbial þa – an action marker? Neuphilologische Mitteilungen 73: 90–96. Enkvist, Nils Erik. 1986. More about the textual functions of the Old English adverbial þa. In: Dieter Kastovsky and Aleksaner Szwedek (eds.), Linguistics across Historical and Geographical Boundaries: In Honour of Jacek Fisiak on the Occasion of his Fiftieth Birthday, Vol. I, 301–309. Berlin/New York: Mouton de Gruyter. Enkvist, Nils Erik and Brita Wa˚rvik. 1987. Old English þa, temporal chains, and narrative structure. In: Anna Giacalone Ramat, Onofrido Carruba, and Guiliano Bernini (eds.), Papers from the 7th International Conference on Historical Linguistics, 221–237. Amsterdam/Philadelphia: John Benjamins. Fischer, Olga, Annette Rosenbach, and Dieter Stein (eds.). 2000. Pathways of Change: Grammaticalization in English. Amsterdam/Philadelphia: John Benjamins. Fries, Udo. 1993. Towards a description of text deixis in Old English. In: Klaus R. Grinda and Claus-Dieter Wetzel (eds.), Anglo-Saxonica: Festschrift fu¨r Hans Schabram zum 65. Geburtstag, 527–540. Munich: Fink. Goffman, Erving. 1974. Frame Analysis: An Essay on the Organization of Experience. New York: Harper and Row. Go¨rlach, Manfred. 2004. Text Types and the History of English. Berlin/New York: Mouton de Gruyter. Gotti, Maurizio, Marina Dossena, and Richard Dury (eds.). 2008. English Historical Linguistics 2006. Selected Papers from the Fourteenth International Conference on English Historical Linguistics (ICEHL 14), Bergamo, 21–25 August 2006. Vol. I: Syntax and Morphology. Amsterdam/ Philadelphia: Benjamins. Grzega, Joachim. 2008. Ha¯l, Hail, Hello, Hi: Greeting in English language history. In: Jucker and Taavitsainen (eds.), 165–193. Healey, Antonette di Paolo (ed.). 2005. Dictionary of Old English Corpus (DOEC). Web Corpus. with John Price Wilkin and Xin Xiang. University of Toronto: Centre for Medieval Studies, Dictionary of Old English Project. http://www.doe.utoronto.ca/pages/pub/web-corpus.html Hiltunen, Risto. 2006. “Eala, geferan and gode wyrhtan”: On interjections in Old English. In: John Walmsley (ed.), Inside Old English. Essays in Honour of Bruce Mitchell, 91–116. Maldon, MA: Blackwell. Hopper, Paul J. 1979. Aspect and foregrounding in discourse. In: Talmy Givo´n (ed.), Discourse and Syntax, 213–241. New York: Academic Press. Jauss, Hans Robert. 1979. The alterity and modernity of medieval literature. Trans. Timothy Bahti. New Literary History 10: 181–227. Jucker, Andreas H. (ed.). 1995. Historical Pragmatics. Pragmatic Developments in the History of English. Amsterdam/Philadelphia: John Benjamins. Jucker, Andreas H. and Irma Taavitsainen. 2000. Diachronic speech act analysis: Insults from flyting to flaming. Journal of Historical Pragmatics 1: 67–95. Jucker, Andreas H. and Irma Taavitsainen. 2008. Speech Acts now and then: Towards a pragmatic history of English. In: Jucker and Taavitsainen (eds.), 1–23. Jucker, Andreas H. and Irma Taavitsainen (eds.). 2008. Speech Acts in the History of English. Amsterdam/Philadelphia: John Benjamins. Kemenade, Ans van and Bettelou Los. 2006. Discourse adverbs and clausal syntax in Old and Middle English. In: van Kemenade and Los (eds.), 224–248. Kemenade, Ans van and Bettelou Los (eds.). 2006. The Handbook of the History of English. Oxford: Blackwell. Kemenade, Ans van, Tanja Milicev, and R. Harald Baayen. 2008. The balance between syntax and discourse in Old English. In: Gotti, Dossena, and Dury (eds.), 3–21. Kim, Taejin. 1992. The Particle þa in the West Saxon Gospels: a Discourse Level Analysis. Bern: Peter Lang. Knappe, Gabriele. 2008. Flyting und die Rhetorik des verbalen Konflikts in der angelsa¨chsischen Literatur. In: Oliver Auge, Felix Biermann, Matthias Mu¨ller, and Dirk Schulze (eds.), Bereit
21 Old English: Pragmatics and discourse zum Konflikt. Strategien und Medien der Konflikterzeugung und Konfliktbewa¨ltigung im europa¨ischen Mittelalter, 32–46. Ostfilden: Jan Thorbecke Verlag. Kohnen, Thomas. 2000. Explicit performatives in Old English: A corpus-based study of directives. Journal of Historical Pragmatics 1: 301–321. Kohnen, Thomas. 2002. Towards a history of English directives. In: Andreas Fischer, Gunnel Tottie, and Hans Martin Lehmann (eds.), Text Types and Corpora: Studies in Honour of Udo Fries, 165–175. Tu¨bingen: Niemeyer. Kohnen, Thomas. 2007a. Connective profiles in the history of English texts: Aspects of orality and literacy. In: Ursula Lenker and Anneli Meurman-Solin (eds.), Clausal Connection in the History of English, 289–308. Amsterdam/Philadelphia: John Benjamins. Kohnen, Thomas. 2007b. Text types and the methodology of diachronic speech act analysis. In: Susan M. Fitzmaurice and Irma Taavitsainen (eds.), Methods in Historical Pragmatics, 139– 166. Berlin/New York: Mouton de Gruyter. Kohnen, Thomas. 2008a. Directives in Old English: Beyond politeness. In: Jucker and Taavitsainen (eds.), 27–44. Kohnen, Thomas. 2008b. Linguistic politeness in Anglo-Saxon England? A study of Old English address terms. Journal of Historical Pragmatics 9: 140–158. Kohnen, Thomas. 2008c. Tracing directives through age and time: Towards a methodology of a corpus-based diachronic speech-act analysis. In: Jucker and Taavitsainen (eds.), 295–310. Lenker, Ursula. 2000. Soþlice and witodlice: Discourse markers in Old English. In: Fischer et al. (eds.), 229–249. Lenker, Ursula. 2010. Argument and Rhetoric: Adverbial Connectors in the History of English. Berlin/New York: Mouton de Gruyter. Lerer, Seth. 1991. Literacy and Power in Anglo-Saxon Literature. Lincoln/London: University of Nebraska Press. Los, Bettelou. 2000. Onginnan/beginnan + to-infinitive in Ælfric. In: Fischer et al. (eds.), 251–274. Rissanen, Matti, Merja Kyto¨, Leena Kahlas-Tarkka, Matti Kilpio¨, Saara Nevanlinna, Irma Taavitsainen, Terttu Nevalainen, and Helena Raumolin-Brunberg. 1991. The Helsinki Corpus of English Texts. In: ICAME Collection of English Language Corpora (CD-ROM), 2nd edn., eds. Knut Hofland, Anne Lindebjerg, and Jørn Thunestvedt, The HIT Centre, University of Bergen, Norway. For manual, see http://khnt.hit.uib.no/icame/manuals/hc/index.htm Sauer, Hans. 2007. Ælfric and emotion. Poetica. An International Journal of Linguistic-Literary Studies ( Japan) 66: 37–52. Sauer, Hans. 2008. Interjection, emotion, grammar, literature. In: Amano Masachiyo, Michiko Ogura, and Masayuki Ohkado (eds.), Historical Englishes in Varieties of Texts and Contexts, 387–403. Bern: Peter Lang. Schiffrin, Deborah. 1987. Discourse Markers. Cambridge: Cambridge University Press. Schlieben-Lange, Brigitte. 1976. Fu¨r eine historische Analyse von Sprechakten. In: Heinrich Weber and Harald Weydt (eds.), Sprachtheorie und Pragmatik, 113–119. Tu¨bingen: Niemeyer. Searle, John R. 1969. Speech Acts: An Essay in the Philosophy of Language. Cambridge: Cambridge University Press. Stanley, Eric G. 2000. Hwæt. In: Jane Roberts and Janet Nelson (eds.), Essays on Anglo-Saxon and Related Themes in Memory of Lynne Grundy, 525–556. London: King’s College, Centre for Late Antique and Medieval Studies. Traugott, Elizabeth Closs. 1991. English speech act verbs: A historical perspective. In: Linda R. Waugh and Stephen Rudy (eds.), New Vistas in Grammar: Invariance and Variation, 387– 406. Amsterdam/Philadelphia: John Benjamins. Wagner, Andreas. 1994. Wie sich Sprechakte historisch vera¨ndern. Vorstudien zu einer Typologie des historischen Wandels von Sprechakten am Beispiel von SEGNEN im Althebra¨ischen und BEKENNEN im Deutschen. In: Dieter W. Halwachs and Irmgard Stu¨tz (eds.), Sprache – Sprechen – Handeln. Akten des 28. Linguistischen Kolloquiums Graz 1993. Bd. 2, 181–187. Tu¨bingen: Niemeyer.
339
340
III Old English Wa˚rvik, Brita. 1995. Peak marking in Old English narrative. In: Brita Wa˚rvik, Sanna-Kaisa Tanskanen, and Risto Hiltunen (eds.), Organization and Discourse. Proceedings from the Turku Conference, 549–558. Turku: University of Turku. Wa˚rvik, Brita. 2011. Connective or “disconnective” discourse marker? Old English þa, multifunctionality and narrative structuring. In: Anneli Meurman-Solin and Ursula Leuker (eds.), Connectives in Synchrony and Diachrony in European Languages. Helsinki: VARIENG. http:// www.helsinki.fi/varieng/journal/volumes/08/wa˚rvik. Wu¨lfing, Johann Ernst. 1894–1901. Die Syntax in den Werken Alfreds des Großen. 3 volumes. Bonn: Haunstein. Yule, George. 1996. Pragmatics. Oxford: Oxford University Press.
Ursula Lenker, Eichsta¨tt (Germany)
22 Old English: Dialects 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
The main Old English dialects Some important people Research on the Old English dialects Some problems for research on Old English dialects Dialects and standard The origin of the Old English dialects The transmission of the Old English dialects Phonology Inflectional morphology Word-formation Vocabulary and word geography The Winchester Vocabulary Wulfstan’s vocabulary The question of a poetic dialect Further development References
Abstract The chapter provides a description of the usual dialectal division of Old English into WestSaxon in the South, Kentish in the South-East, and Anglian in the Midlands and the North, with a further subdivision of Anglian into Mercian and Northumbrian. Some of the AngloSaxon personalities that are connected with the various dialects are first introduced. A sketch of research on the OE dialects along with some problems for research follows. The chapter then deals briefly with the question of dialects and standard language and with the origin and transmission of the OE dialects. The main dialectal differences in phonology, inflexional morphology, word-formation, vocabulary, and word-geography are outlined, with special attention paid to the (early) runic inscriptions, the Winchester Vocabulary, and Wulfstan’s Vocabulary. As far as the transmission allows us to judge, there were no differences in syntax. The chapter ends with an exploration of the question of a poetic dialect and of the role of the OE dialects in the further development of the English language. Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 340–361
22 Old English: Dialects
341
1 The main Old English dialects Old English was split up into several dialects from the beginning. Usually three main Old English dialects are distinguished, namely, West-Saxon (WSax.) in the South, Kentish (Kent.) in the South-East (Kent and neighbouring areas), and Anglian (Angl.) in the Midlands and the North. Anglian can be subdivided into Mercian (Merc.; roughly the Midlands) and Northumbrian (Nhb.; roughly north of the Humber); see Map 22.1 below. The names of the dialects are partly connected with the members of the Germanic tribes that according to Bede’s Historia ecclesiastica I.15 (Colgrave and Mynors [eds.] 1969) came in 449 CE from what is today Northern Germany and Southern
NORTHHUMBRIA
M E R W A L E S
C I A
KENT WESSEX Cornwall
Map 22.1: The OE dialect boundaries based on the political boundaries of c.825 (map based on Sievers-Brunner 1965)
342
III Old English Denmark. They sailed across the North Sea and conquered Britain; see Map 22.2 below. West-Saxon is connected with a part of the Saxons (there were also South-Saxons and East-Saxons, who, however, do not play much of a role when talking about OE dialects); Anglian is connected with the Angles; and Kentish is connected with the Jutes, who settled in Kent and also the Isle of Wight. How far there is a Jutish substratum in the WSax. dialect of Winchester is disputed (see, e.g., Seebold 1990; Sauer 1992: 322–323). Possibly Frisians also played a role in the invasion and conquest of Britain, but they do not seem to be specifically connected with any one of the OE dialects. Old English and Old Frisian as such are closely related, however, and there are a number of common Anglo-Frisian features. Some sort of standard developed only in the later OE period on the basis of Late West-Saxon (LWSax.); see Section 5 below. The political dominance moved from North to South (with some overlap): the Northumbrians were dominant in the 7th century (c.625–675), the Mercians in the 8th century (c.650–825), and the WestSaxons from the 9th century to the Norman Conquest (c.800–1066). The Kentish were never politically dominant (see Toon 1992: 416), although Canterbury was important as the seat of an archbishop – there were only two archbishoprics in Anglo-Saxon England, namely Canterbury (from 597 onwards) and York (from 735 onwards).
JU
LES
Ba
ltic
Se
a
S
I R
er
F
r
N S S A X O
be Od
se
S
El
We
Ems
I
N
S
ANG
Nor th Sea
A
TE
Rhin e
5°
10°
15°
Map 22.2: The homes of the Anglo-Saxons (taken from Baugh and Cable 2002: 48)
20°
22 Old English: Dialects
2 Some important people Although not only most OE speakers, but also many literary authors and scribes remain anonymous, we know the names of some individuals that are connected with certain dialects. Apart from King Alfred, all of them were (or became) clerics (e.g., monks, priests, bishops). We list them here in alphabetical order (for more details see, e.g., Lapidge et al. [eds.] 1999): – Ælfric of Eynsham (c.950–c.1010): pupil of Æthelwold at Winchester, later monk at Cerne Abbas, and finally abbot of Eynsham; the most important LWSax. prose author and representative of the Winchester Vocabulary. – Æthelwold (c.904/909–984): abbot of Abingdon, later bishop of Winchester; one of the main proponents of the Benedictine Reform; teacher of Ælfric and founder of the Winchester Vocabulary. – Aldred (fl.950–970): priest at Chester-le-Street; added the Late Northumbrian gloss to the Lindisfarne Gospels and the Durham Ritual. – Alfred (849–899): king of Wessex (871–899); defeated the Vikings; also attempted an educational reform and assembled a group of learned helpers. He translated a number of Latin texts into early West-Saxon; his helpers also translated or compiled some texts – Alfred has therefore been called “the father of English prose”. – Bede or Beda Venerabilis (c.673–735): monk at Monkwearmouth-Jarrow; the most important Anglo-Saxon theologian and historian; wrote mainly in Latin, but his Death-Song originally in the early Northumbrian dialect. – Byrhtferth of Ramsey (c.970–c.1020): monk at Ramsey and prolific author, mainly in Latin, but wrote his Enchiridion (or Handbook) partly in Late West-Saxon. – Cædmon (fl. c.660–c. 680): illiterate cow-herd and later monk at Whitby; model of an oral poet whose poetry was written down by others; also the first English poet who is known by name; composed his Hymn originally in the early Northumbrian dialect. – Cynewulf (perhaps around 900): a learned Mercian poet who signed his poems with runic letters. – Farmon (second half of 10th century): a priest who glossed part of the Rushworth Gospels (Rushworth1) in Mercian. – Owun (second half of 10th century): a scribe, and probably also a cleric, who glossed part of the Rushworth Gospels (Rushworth2) in Northumbrian, closely following Aldred’s gloss to the Lindisfarne Gospels. – Wærferth of Worcester (c.840–915): bishop of Worcester (872–915); one of King Alfred’s learned helpers; translated the Dialogues by Pope Gregory the Great into Old English (Anglian or, more specifically, Mercian). – Wulfstan the homilist (c.960–1023): bishop of London (996–1002); archbishop of York (1002–1023) and, for a long time simultaneously, bishop of Worcester (1002– 1016); besides Ælfric one of the most important LWSax. prose authors (but not connected to the Winchester Vocabulary).
343
344
III Old English
3 Research on the Old English dialects The differences between the OE dialects were mainly recognized by 19th century scholars, e.g. by Sweet (1887) (on the history of research, cf. Campbell 1959: 4; Toon 1992: 434–437). There are many studies devoted to specific problems, but only a few comprehensive surveys. Phonology and inflectional morphology on the one hand and vocabulary on the other are mostly treated separately. The major grammars and handbooks, such as Campbell (1959), Hogg (1992a), and Sievers-Brunner (1965), which goes back to Sievers (1882), deal with dialectal differences in phonology (and inflection) but not in vocabulary. Even in the first volume of the Cambridge History of the English Language (Hogg [ed.] 1992b), the chapter by Toon on “Old English dialects” concentrates on the historical and social background and on phonological dialect markers, whereas dialect vocabulary is discussed by Kastovsky in the chapter on “Semantics and Vocabulary”, and the question of a poetic dialect is treated by Godden in the chapter on “Literary Language”.
4 Some problems for research on Old English dialects Research on the Old English dialects is fraught with a number of problems. (See also Section 11.2 and Section 14 below.) Written language: what we have is only what was written down, usually for a specific purpose, e.g. for religious instruction of various kinds. It is, of course, even more difficult to trace the spoken language from the written documents. Gaps in the material: the chronological as well as the geographical distribution of the material is very uneven. Apart from a few early runic inscriptions (see Waxenberger 2010: esp. 128–167; Park and Waxenberger 2011: 10) and a few special cases such as the laws of the early Kentish kings (see Section 7.4 below), the written transmission of Old English began only around CE 700. There are, however, not many documents from the 8th and 9th centuries: Ker (1957: xv) lists only eight manuscripts from this period, and 21 from the 10th century. Early personal and place names are also recorded in Latin texts, e.g. Bede’s 731 Historia ecclesiastica (Colgrave and Mynors [eds.] 1969). The bulk of manuscripts containing Old English (more than 130) dates from the 11th century, i.e., from the period before or just after the Norman Conquest (see Ker 1957: xv–xviii). Moreover, throughout the OE period the distribution of the evidence for the dialects is also very uneven (see below). No isoglosses or exact dialect boundaries can therefore be drawn. Some centers of manuscript production are known (see, e.g., Gneuss 2001; Ker 1957), such as Abingdon, Bath, Bury St. Edmunds, Canterbury (especially Christ Church, i.e., the cathedral), Chester-le-Street, Durham, Exeter, Glastonbury, Hereford, Lichfield, Malmesbury, Rochester, Winchester, Worcester, York (strikingly, there seem to be very few manuscripts from London), but not all manuscripts that were copied there show dialect features of the area; in particular, many manuscripts copied at Canterbury do not show Kentish features. Dialectal adaptation: if scribes had to copy an OE text written in a dialect different from their own, they often adapted it to their own dialect by changing the spelling and the vocabulary. Such changes can sometimes be seen relatively easily in texts which have been transmitted in several manuscripts with adaptation to different dialects, e.g. in
22 Old English: Dialects Cædmon’s Hymn and Bede’s Death Song (see Section 7.3 below), also in Wærferth of Worcester’s originally Mercian translation of the Dialogues of Gregory the Great, or the originally Mercian translation of Bede’s Historia ecclesiastica, both connected with King Alfred’s court (cf., e.g., Schabram 1965: 42–48). Changes in dialectal forms are usually difficult to notice or to prove in texts which are transmitted in just one manuscript, because there is nothing to compare them with – this is particularly true of the majority of the Old English poetry, but also of many prose texts. Usually the direction of adaptation was from Anglian (Mercian or Northumbrian) originals to (Late) West-Saxon adaptations, e.g. as with much of the OE poetry (including Cædmon and Bede), rarely the other way round. But we should not automatically conclude from this that Angl. texts are normally early and WSax. texts are normally late; counterexamples are, e.g., the Northumbrian glosses to the Lindisfarne Gospels, which date from the second half of the 10th century. Mixed texts: whether dialectally mixed texts are always the result of the WestSaxonization of originally Angl. texts as just described, is not certain. In some cases there may have been a dialectal mixture from the beginning. In other cases, especially with glosses and glossaries, which were sometimes compiled over a longer period of time and by glossators with different dialectal backgrounds, a mixture may have accordingly accumulated in the course of time. Mobility of authors and scribes: some authors and scribes moved around, which means that their dialects do not necessarily represent the dialect of the place where they were working. Thus Wulfstan, who wrote in Late West-Saxon, was originally bishop of London, but later bishop of Worcester (in the West Midlands, i.e., the Mercian dialect area) and archbishop of York (in the Northumbrian dialect area).
5 Dialects and standard Dialects are usually seen as opposed to the standard language. Linguistic lay people often think that dialects are a sort of deviant (or even corrupt) form of the standard language. But dialects normally exist earlier than the standard language, and the standard language usually develops from the dialect of that region that is, for political, economic, or cultural reasons (or a combination of them), more powerful than other regions with their dialects (there is a saying amongst linguists that a standard language is a dialect with an army and a navy). This is certainly true for Old English: the dialects existed first, and a sort of standard developed only much later (see further Kornexl, Chapter 24). Due to the political dominance of the West-Saxon kingdom from the 9th century onwards, it was natural that West-Saxon eventually developed into a kind of standard language (cf., e.g., Godden 1992: 518–520). The extent to which the Early West Saxon of King Alfred (died 899) and some of his collaborators was already a standard is, however, difficult to say for several reasons. Few manuscripts from Alfred’s time survive (see Ker 1957: xv), and in these, spelling is sometimes inconsistent. Alfred had both West-Saxon and Mercian helpers (among the latter were Wærferth of Worcester and the anonymous translator of the Old English Bede) and even helpers from the Continent, which shows that Anglian or Mercian was certainly tolerated at Alfred’s court. In the second half of the 10th century, Late West-Saxon apparently developed into a kind of standard language (OE manuscripts from the first half of the 10th century are
345
346
III Old English rare). This can be seen from the large number of manuscripts from the second half of the 10th and from the 11th century which have WSax. features. As mentioned above, originally Anglian texts were often West-Saxonized, including the bulk of OE poetry. Ælfric in particular seems to have been very concerned about correct usage. How far this standard extended is difficult to say, however. As can be seen from Aldred’s glosses, Northumbrian was not affected by West-Saxon even in the second half of the 10th century. Canterbury, on the other hand, had been a meeting place of Angl. and WSax. influences for a long time and seems to have been mainly under WSax. influence in the 10th and 11th centuries (see below). And even within Late West-Saxon there were differences, perhaps representing different sub-dialects; see especially Sections 12 and 13. What we have, moreover, is only the written standard; we do not know how far there was a spoken standard as well. Usually the written standard emerges earlier and is more widespread than the spoken standard. In many countries today, including England and Germany, there is a written standard language, but there are numerous regional accents and dialects. The situation will likely have been similar in Late Anglo-Saxon England.
6 The origin of the Old English dialects The question to be addressed here is the extent to which the differences between the OE dialects developed in England after the settlement of the Anglo-Saxons (i.e., after c.450) and the extent to which the Anglo-Saxons, who came from different areas (i.e., Angeln, Saxony, Jutland; see Section 1. above and Map 22.2), brought dialectal differences from the Continent with them. This question is difficult to answer (see, e.g., Nielsen 1985). One of the problems is that the written transmission of Old English began only around 700 (see Section 4 above), and the transmission of the Continental Germanic dialects still later. The early OE runic inscriptions can shed some light on the sound changes in Pre-OE (e.g., fronting of Grmc. a/a¯; see Waxenberger 2010: 128–155; Park and Waxenberger 2011: 9). According to Campbell (1959: 110) the distinction between WSax. æ¯ and Anglian e¯ (see below) was most probably brought from the Continent, whereas all other phonological differences arose in England. As far as vocabulary is concerned, Korhammer (1980) points out that a few differences in vocabulary which exist between the MSS of the Old Saxon poem Heliand (first half of the 9th century) correspond to differences between Anglian and West Saxon. Fragment S of the Heliand has words which are paralleled in Anglian, and MSS CM have words which are paralleled in the common OE vocabulary or in West Saxon. All pairs or groups of synonyms in question are function words (adverbs or conjunctions), see Table 22.1 below. The conclusion Korhammer draws from this lexical evidence is that some differences between Anglian and West Saxon go back to varieties within Old Saxon. Some of these correspondences are more striking than others, however, and Nielsen (1991: 244–249, 260–261) doubts that the parallels are weighty enough to justify the postulation of a dialectal difference within Old Saxon which then lived on as a dialectal difference within Old English. Moreover, he finds no phonological or morphological correspondences of the fragment S with Anglian. So the debate about the origin of the OE dialects will probably continue.
22 Old English: Dialects
347
Table 22.1: Vocabulary differences within Old Saxon and within Old English Old Saxon Heliand: Fragment S
Old Saxon Heliand: MSS CM
OE Anglian
OE Common Old English and West Saxon
tulgo ‘very’
switho, swiðo tegegnes bi hui
tulge, tylige, etc. togeaegn for hwon ‘why’ bi hwon, etc. ‘where from’
swiðe
tigene ‘against’ bi huon, etc. ‘why’
togeagnes, togenes for hwi/hwy to hwi/hwy
7 The transmission of the Old English dialects As mentioned above, the transmission of the OE dialects is very uneven. Late WestSaxon is attested best and transmitted in many texts and manuscripts, whereas Kentish has the weakest attestation, and many originally Anglian texts survive only in LWSax. adaptations, for example the bulk of OE poetry (see, e.g., Campbell 1959: 4–11; Hogg 1992a: 3–8; Sievers-Brunner 1965: 2–11; Toon 1992: 422–428).
7.1 West-Saxon This dialect is often divided into Early and Late West-Saxon. Early West-Saxon is basically represented by the writings and translations of King Alfred (849–899), and some other works that probably originated in his circle and were written, compiled, or translated by his West-Saxon helpers. Many scholars accept that the OE translations of Pope Gregory’s Cura Pastoralis, of the prose portion of the Paris Psalter, of Boethius’s Consolatio Philosophiae, and of St. Augustine’s Soliloquies, as well as Alfred’s law code, were made by Alfred, whereas the OE (WSax.) translation of Orosius’s Historia Adversus Paganos was made by one of Alfred’s West-Saxon helpers and the original (WSax.) version of the Anglo-Saxon Chronicle was also compiled in his circle. (see, e.g., Bately 2009; Frantzen 1986; and Godden 2007 for a dissenting voice). But only the OE versions of the Cura Pastoralis, the Boethius, and the Orosius as well as the Anglo-Saxon Chronicle and the laws survive in more or less contemporary manuscripts, whereas the Soliloquies are transmitted in a manuscript from around 1150. Late West-Saxon, on the other hand, is by far the best represented OE dialect, and probably even served as a sort of standard language (at least as a written standard). It began around 950 and was dominant throughout the rest of the Anglo-Saxon period. Many texts are anonymous, e.g. the West-Saxon Gospels, but four named authors who represent Late West-Saxon stand out: Abbot Ælfric, the most productive Old English prose writer; Ælfric’s teacher, bishop Æthelwold of Winchester; Wulfstan, first bishop of London and later bishop of Worcester and archbishop of York, and Byrhtferth of Ramsey (see Section 2).
348
III Old English
7.2 Mercian The main Mercian document, and also the earliest, is the OE gloss to the Vespasian Psalter, which was written around 850 (see Ker 1957: no. 203); there are also some originally Mercian charters (on charters, see Sawyer 1968; Kelly 1999). The Vespasian Psalter Gloss has been connected with a Mercian literary language; the latter has also been postulated for the Life of St Chad, probably an OE text, but transmitted only in a 12th century MS (see Vleeskruyer 1953). A continuation of this can be seen in the Early Middle English, so-called AB-language of the Ancrene Riwle (or Ancrene Wisse) and the Katherine Group (around 1200). ´ pinal-Erfurt Glossary Other Mercian texts, at least in their original form, are the E (which originated around 700; cf., e.g., Ker 1957: no. 114) and the Corpus Glossary (c.800; cf., e.g., Ker 1957: no. 36). Wærferth of Worcester’s translation of Gregory’s Dialogues and the OE translation of Bede’s Historia ecclesiastica were originally Mercian, although both are connected with King Alfred’s West-Saxon court (before 900). A later Mercian text is the gloss called Rushworth1 in the Rushworth Gospels (cf. Ker 1957: no. 292) written by Farmon.
7.3 Northumbrian Early Northumbrian is attested by only a few short manuscript texts which were originally composed in the late 7th or early 8th centuries, namely, Cædmon’s Hymn (which is transmitted in several manuscripts of Bede’s Historia ecclesiastica), Bede’s Death Song, and the Leiden Riddle. Cædmon’s Hymn and Bede’s Death Song are also transmitted in West-Saxon versions. The corpus of Old English runic inscriptions, although relatively small (96 inscriptions altogether; see Waxenberger 2010: 16–18 and Map 22.3), provides valuable information on early Northumbrian as well. Two of the most prominent runic texts are on the Franks Casket (also called Auzon Casket; cf., e.g., Becker 1973; Waxenberger 2010: 28–38; 554–575; Park and Waxenberger 2011: 16–21) and on the Ruthwell Cross (Dumfrieshire; see, e.g., Waxenberger 2010: 92–99). The latter holds a short version of the Dream of the Rood, a poem which also exists in a later, longer, and mainly West-Saxon version in the Vercelli Book. In the cases of Cædmon’s Hymn, Bede’s Death Song, and the Dream of the Rood, Early Northumbrian and Late West-Saxon versions can be directly compared. Late Northumbrian is attested in the second half of the 10th century by three long interlinear glosses, namely the gloss to the Lindisfarne Gospels, the Northumbrian part of the gloss to the Rushworth Gospels (so-called Rushworth2) written by Owun, and the gloss to the Durham Ritual. The glosses in the Lindisfarne Gospels and also probably in the Durham Ritual were written by Aldred.
7.4 Kentish Kentish is sparsely represented. There are some Kentish charters from the 9th century as well as the Kentish glosses and two poems (the Kentish Psalm and the Kentish Hymn) in manuscript Cotton Vespasian D.VI from the middle of the 10th century (see Ker 1957: no. 207; Kalbhen 2003). Some Kentish features exist in the texts of MS Cotton Tiberius A. iii (Ker 1957: no. 186). One special case is the set of laws of
22 Old English: Dialects
349
Lindisfarne
Cramond Alnmouth
Coquet Island Falstone Bewcastle
Ruthwell Mote of Mark St. Ninian’s Cave
Whithorn
Maughold
Monkwearmouth
Chester-le-Street Wheatley Hill Durham Hartlepool
Kingmoor
NORTHHUMBRIA
Whitby Hackness
Great Urswick
West Hestlerton
Malton Lancaster Collingham York Bingley Bramhan Moor Leeds Crowle Welbeck Hill Kirkheaton Thornhill Cleatham Scutterthorpe Morton Bakewell Overchurch Leek Lovedon Hill
M
E
Llysfaen
R
Derby
W A L E S
C
Heacham
I
Wardley
A
Keswick Brandon
Wakerly
Long Buckby
Gayton Thorpe Sponge Hill
Caistor-by-Norwich
Undley Schropham
Blythburg
Worcester
London
Watchfield
Sarre Orpington Boardley
Gilton
Sandwich Dover
WESSEX
Southampton Selsey Chessel Down Carisbrook
Cornwall
Map 22.3: Findspots of the Old English runic inscriptions
the early Kentish kings (especially Æthelberht). They were originally written down shortly after 600 (i.e., shortly after the conversion), but are only preserved in a MS from the 12th century (the Textus Roffensis; cf. Ker 1957: no. 373).
8 Phonology In the following section, the main Old English phonological dialect differences are outlined in approximate chronological order. These differences are due to sound changes that occurred in the dialects. The chronology of these changes is only relative and not
350
III Old English absolute; i.e., some sound-changes are clearly earlier than others, but often the soundchanges cannot be dated to a specific period. The changes mentioned in Sections 1–7 are generally called prehistoric; that is, they must have taken place between c.450 and c.700, i.e., before the time of the earliest written manuscripts. The changes mentioned in Sections 8–11, on the other hand, can be seen in the manuscript texts. Many words were affected by two or more sound-changes in sequence, some of which were dialectally significant and others not. OE phonology is a complex area; for more details see the handbooks and historical grammars, e.g., by Campbell (1959); Hogg (1992a); Luick (1921); Sievers-Brunner (1965); Toon (1992: 430–451); Waxenberger (2010) (see further Murray, Chapter 17). We have marked vowel length systematically in this section but not in the others unless it was absolutely necessary. Some sound changes are restricted to a specific dialect, whereas others are simply more frequent in one dialect; i.e., in the latter cases we have to make do with tendencies rather than rules – and there is, of course, the problem of dialectally mixed texts, see above. (1) Development of WGrmc. *a¯: WGrmc. *a¯ developed into e¯ in Anglian and Kentish, but into æ¯ in West-Saxon. This affected native words (where it goes back to IE e¯ > Grmc. æ¯ [so-called e¯1] > WGrmc. a¯) as well as early Lt. loan-words, e.g. Angl. e¯ton – WSax. æ¯ton (cf. Lt. edimus; Ger. aßen; ModE ate); Angl. de¯d – WSax. dæ¯d (cf. Ger. Tat; ModE deed); Angl. se¯d – WSax. sæ¯d (cf. Ger. Saat; ModE seed); Lt. (via) stra¯ta > Angl. stre¯t – WSax. stræ¯t (cf. Ger. Straße; ModE street). (2) Grmc. *a before nasal: Grmc. *a before nasals developed into the nasalized allophone OE [a˚] (this symbol is used here to mark the difference in quality between the long nasalized [*a˜ː] > OE [o¯] and the short [a˚]) which was written or . Angl. texts show mostly . Both spellings occur in Early Northumbrian; in the 10th century Northumbrian texts is practically universal. ´ pinal-Erfurt Glossary and Corpus Glossary, The oldest Mercian glossaries, E have both and . The Mercian Vespasian Psalter has almost exclusively . In early WSax. texts both and were used, whereas in LWSax. (Ælfric; West Saxon Gospels) was almost always used. In 9th century Kent. manuscripts there was mostly but predominantly in the 10th century texts (cf., e.g., Campbell 1959: 51–52; Sievers-Brunner 1965: 52; Waxenberger 2010: 276–279, 371–372). (3) Breaking (or fracture) and retraction: the short palatal vowels æ, e, i before [x] , r, l + consonant or simple [x] were diphthongized to short æu, eu, iu, which then developed to short ea, eo, io > eo. Breaking before [x] (+ consonant) and before r + consonant took place in all dialects, whereas breaking of æ before l + consonant only took place in West Saxon and Kentish; in Anglian, it was usually retracted to al + consonant. In Anglian, the other diphthongs were also often monophthongized, if a velar vowel followed (Angl. smoothing; see no. 8 below). Examples are: WSax./Kent. eahta – Angl. æhta (Angl. smoothing) (Ger. acht; ModE eight); WSax./Kent. bearn ‘child’ – Angl. barn (retraction); WSax./Kent. heard – Angl. hard (retraction) (Ger. hart, ModE hard); WSax./Kent. eall, feallan, wealdan – Angl. all, fallan, waldan (Ger. all(e), fallen, walten; ModE all, fall, wield); WSax./Kent. ceald – Angl. cald (Ger. kalt; ModE cold); WSax./Kent.
22 Old English: Dialects
(4)
(5)
(6)
(7)
feohtan – Angl. fehtan (Angl. smoothing) (Ger. fechten; ModE fight); early OE *hiordi > WSax. hierde (i-umlaut) – Nhb. hiorde; Merc. heorde (as io > eo in Merc./WSax.) (Ger. Hirte; ModE [shep]herd, herds[man]). But compare: OE helpan, meltan without breaking (ModE help, melt). In LWSax. eo, io by breaking were later monophthongized to ie > i before [xs] , [xt] , e.g. OE cneoht > cnieht > cniht (Ger. Knecht; ModE knight); see also 8 below. Second fronting: this is apparently mainly limited to the Mercian dialect of the Vespasian Psalter Gloss. Grmc. short a developed generally to æ in Pre-Old English (Anglo-Frisian fronting), but was retracted (restored) to a before velar vowels, e.g. Grmc. *daga- > OE SG dæg, but PL dagas. In the Mercian dialect of the Vespasian Psalter, however, æ was fronted to e, and a to æ, e.g. SG deg, PL dægas (ModE day, days); OE fæder – Vespasian Psalter feder (ModE father). Palatal diphthongization: after initial palatal consonants (/tʃ/ , /j/ g˙, /ʃ/ ), primary palatal vowels (short and long æ/æ¯, e/e¯), i.e., palatal vowels not due to i-mutation, were diphthongized, yielding ea/e¯a, ie/ı¯e. Therefore, æ/æ¯ developed to ea/e¯a and e/e¯ became ie/ı¯e (short and long). Palatal diphthongization was most frequent in West-Saxon and occurred also in Late Northumbrian, but did not take place in Mercian, Kentish, and runic early Northumbrian. It is thus noticeable that this sound-change was shared by two distant dialects, but not by neighbouring dialects. Examples are: WSax. g˙iefan [jɪəvan] – Merc./Kent. g˙efan (Ger. geben; ModE give); WSax. g˙ietan [jɪətan] – Merc./Kent. g˙eta(n) (ModE get); WSax. sc˙e¯ap – Mercian/Kent. sc˙e¯p (Ger. Schaf; ModE sheep); cf. also WSax. g˙eong; c˙easter; g˙e¯ar (ModE young, -chester, year). i-mutation (i-umlaut): apart from the palatal vowels e and i, all vowels and diphthongs (a, a¯, æ, æ¯, o, o¯, u, u¯, ea/e¯a, io/ı¯o) were fronted (or raised), if the following syllable contained an i or j; the i, j which caused the i-mutation was then usually lost (or in some cases lowered to e). Thus: a > æ (a + nasal > e); o/o¯ > œ/œ̅ /œ(ː)/ (often retained in Anglian) > e/e¯ (in WSax.); u/u¯ > y/y¯ /y(ː)/; ea/e¯a > WSax. ie/ı¯e (> i/ı¯, y/y¯), Angl./ Kent. e/e¯; io/ı¯o > WSax. ie/ı¯e (> i/ı¯, y/y¯), Angl./Kent. io/ı¯o. Although i-mutation was generally carried through in Old English, there were dialectal differences, partly due to different preceding sound changes (see above). Common OE examples are, e.g. *sand-jan > OE sendan (Ger. senden; ModE send); dohtor, DAT SG *dohtri > dœhter > WSax. dehter (Ger. Tochter; ModE daughter); *kuning- > OE cyning (Ger. Ko¨nig; ModE king). Dialectal differences occur, for example, in: *so¯k-jan > Angl. sœ̅ can – WSax. se¯can (Ger. suchen; ModE seek, be-seech); COMPR of WSax. eald – Angl. ald: WSax. *eald-ira > ieldra (> yldra) – Angl. *ald-ira > ældra (Ger. a¨lter; ModE older); WSax. clæ¯ne – Kent. cle¯ne (ModE clean); *mahti-, *nahti- > WSax. *meahti-, *neahti- (breaking) > mieht, nieht > miht, niht – Angl. *mæhti-, *næhti(smoothing) > meht, neht (Ger. Macht, Nacht; ModE might, night); Lt. ca¯seus > WGrmc. *ka¯si (Ger. Ka¨se) > *cæ¯si > WSax. *ce¯asi (palatal diphthongization) > cı¯ese (i-mutation) – Angl./Kent. *ce¯si (i-mutation) > ce¯se (ModE cheese); Grmc. *hauz-jan > early OE *he¯ar-jan > WSax. hı¯eran – Angl./Kent. he¯ran (Gr ho¨ren; ModE hear). Back mutation (velar umlaut): around the 8th century, the short front (palatal) vowels æ, e, i were diphthongized before simple consonants and a back (velar) vowel (u, o, a) in the following syllable: æ > ea, e > eo, i > io > eo. The effect was the same as with breaking, but the cause was a different one. Contrary to breaking,
351
352
III Old English
(8)
(9)
(10)
(11)
back mutation was most frequent in Anglian and Kentish, but rarer in West-Saxon. In West-Saxon, it occurred only before labials and liquids (especially l, r), in the other dialects, before all consonants. Examples are: OE eofor ‘boar’ (Ger. Eber); WSax./Kent. setol – Angl. seotul, seatol ‘seat’; Kent. spreocan – WSax./Angl. sprecan (Ger. sprechen; ModE speak). Anglian smoothing: in Anglian, the diphthongs ea/e¯a, eo/e¯o, io/ı¯o (which had often been brought about by breaking) were monophthongized to æ/æ¯, e/e¯, i/ı¯ before the velar consonants /k/, [ɣ], [x] (simple or in combination with r or l). Examples are: WSax. be¯acon – Angl. be¯con (ModE beacon); WSax./Kent. feohtan – Angl. fehtan (Ger. fechten; ModE fight); WSax./Kent. se¯oc – Angl. se¯c (ModE sick); WSax./Kent. le¯oht (< early WSax. lı¯oht) – Angl. lı¯ht ‘light’. Late West-Saxon smoothing: in Late West-Saxon, ea/e¯a were monophthongized to e/e¯ after initial palatal [tʃ] c˙, [c¸] , /k/ , /j/ g˙, /ʃ/ , e.g. WSax. c˙ealf > LWSax. c˙elf (Ger. Kalb; ModE calf ); WSax. g˙e¯ar (< æ¯; see no. 1 above) > LWSax. g˙e¯r – Angl. g˙e¯r (< *e¯) (Ger. Jahr; ModE year). Thus Angl. and LWSax. forms partly look the same. Other Late West-Saxon changes: weor-, (Early WSax. wier-, wir- >) wyr-, wor- > wur-, e.g. sweord > swurd ‘sword’; weorðan > wurðen ‘to become’; weorold > wuruld ‘world’; wiersa > wyrsa > wursa (ModE worse); furthermore sel- > syl-, sil-, e.g. sylf ‘self’, syllan ‘to give’. Development of OE y/y¯ in Kent.: OE y/y¯ (which had developed due to i-umlaut, see 6 above) was unrounded and lowered to e/e¯ in Kentish: e.g. WSax./Angl. yfel – Kent. efel (Gr u¨bel; ModE evil); WSax./Angl. myrig˙ – Kent. meri(g˙) (ModE merry); WSax./Angl. syn(n) – Kent. sen(n) (Ger. Su¨nde; ModE sin).
In sum, West-Saxon is characterized by the development of WGrmc. a¯ to æ¯; by extensive use of breaking; and by palatal diphthongization. Anglian, on the other hand, is characterized by the development of WGrmc. a¯ to e¯, by retraction (of æ to a), smoothing, and back-mutation. The Mercian Vespasian Psalter Gloss is characterized by second fronting. Kentish lost æ/æ¯, œ/œ̅ , y/y¯ (æ > e; æ¯ > e¯; y > e; y¯ > e¯); as a result there was a predominance of e/e¯ in Kent.
9 Inflectional morphology Compared to phonology, there were not as many dialectal differences in inflexional morphology. Differences concern mainly the verbs. There are many specific differences concerning single verbs (as well as nouns, pronouns, etc.), but these are often difficult to systematize; for details see, e.g., Campbell (1959), Sievers-Brunner (1965). Some of the more systematic differences are: (1) personal pronouns: in the accusative of the 1/2P SG and PL, WSax. has me, þe, us, eow, whereas Angl. has mec, þec, usig, eowic (ModE me, thee, us, you). (2) strong verbs: in WSax. but not in Angl, strong verbs have syncope (and vowel change by i-umlaut and by older e > i; in Angl. the i-umlaut and the change e > i is leveled away) in the 2/3P SG IND PRES; e.g. beran ‘bear, carry’: WSax. bierst, bierþ – Angl. berest, bereþ; c˙e¯osan (> ModE choose): WSax. c˙iest, c˙iesþ – Angl.
22 Old English: Dialects c˙eosest, c˙eoseþ; *sehan [sexan] > se¯on: WSax. 2P siehst, siehþ – Angl. sehst, sehþ (Ger. sehen, 2/3P SG IND PRES siehst, sieht, ModE see); weorþan ‘become’ (Ger. werden): WSax. wierþ – Angl. weorþeþ (Ger. wirst, wird). (3) Northumbrian sometimes has the 2/3P SG and PL in – (e)s; for the 3P SG this is the form that ultimately prevailed in ModE: WSax./Mercian bindest 2P SG, bindeþ 3P SG, bindaþ PL – Nhb. bindes (SG), bindas (PL) (ModE binds 3P SG); see, e.g., Sievers-Brunner (1965: 271–275); Campbell (1959: 299–301). (4) weak verbs class 2: they form their past in WSax. usually in -ode, e.g., lufode, in Angl. and Kent. usually in -ade, lufade (ModE loved).
10 Word-formation There were also differences in word-formation between Anglian and West-Saxon, especially with suffix formations (cf. Kastovsky 1992: 349–351): (1) verbal nouns in -ness (-nis, -nys) were typically formed with the stem of the verb in Angl. and sometimes also in Early WSax., whereas in LWSax. they were mostly formed with the past participle, e.g. forgiefan (> ModE forgive): Angl. forgyfnys, LWSax. forgifennis. A comprehensive study of these formations has apparently not been made, however (see Sauer 1978: 241, with reference to earlier literature). But Gneuss (1955: 161–162) finds relatively many derivations with -nis from the past participle in the (Mercian) Vespasian Psalter. (2) feminine agent nouns were formed with -estre in WSax., but with -icge in Angl., e.g. WSax. hearpestre ‘female harpist’, but Angl. dryicge ‘sorceress’; see Schabram (1970). The WSax. suffix survives in ModE as -ster (gangster), whereas -icge died out. (3) the adjective-forming suffix (or rather suffixoid) derived from the verb beran (> to bear) with the meaning ‘bringing about, carrying’ (for Lt. -fer, e.g. lucifer) normally has the form -berend(e) in Angl., but -bære in WSax., e.g. Angl. deað-berend ‘deadly, mortal’ (lit. ‘death-bringing’), but WSax. æppelbære ‘carrying apples’; see von Lindheim (1972). (4) on the status of Aldred’s Northumbrian loan-formations (loan-translations) based on Latin models, see Section 11.5 below.
11 Vocabulary and word geography 11.1 Principles and history of research Word geography is concerned with the regional distribution of words, more precisely, of synonyms. Most words belonged to the common OE vocabulary, but some were dialectally, i.e., geographically, restricted. Since much of the OE material consists of glosses to Latin words and of translations of Latin texts, words which are used to render the same Latin word or the same meaning of a Latin word can usually be regarded as synonyms. Research on OE word-geography and on dialect vocabulary generally began later and has been less intensive than research on phonology and inflexional morphology. The following section is partly based on Sauer (1992). After the pioneering study by Jordan (1906), a fresh start was made by Schabram
353
354
III Old English (1965) and Wenisch (1979). Kitson (1995) looked at the regional vocabulary used in charter boundaries. Schabram showed that for ‘proud, pride’, Lt. superbus, superbia, Anglian used oferhygdig, oferhygd, whereas West-Saxon used ofermodig, ofermod, ofermedla, etc. and (as was shown later) the LWSax. Winchester Group used modig, modignes (modig occurs in the Angl. poetry, but in the positive sense ‘high-spirited’). The ModE words proud, pride are among the very few French loan-words in Old English. They were borrowed very late (Late OE prud, pryto). Schabram also showed that if originally Anglian texts were West-Saxonized, the original dialect vocabulary was often better preserved than phonologic features. The replacement of Angl. words by their WSax. or common OE synonyms shows moreover that the WSax. revisers must have known the meaning of the Angl. words.
11.2 Problems of research on word-geography and dialect vocabulary The dialectal character of words can change over time. A number of words occur in Angl. and in Early WSax. texts but not in Late West-Saxon. These words were common OE words until the end of the 9th century, but were then dropped from WSax. usage and continued as Angl. dialect words in the 10th and 11th centuries. They died out eventually; their restriction to Anglian can be regarded as a step towards their final obsolescence. Some examples are: blinnan ‘stop’ for Lt. sinere, cessare; carcern ‘prison’ for Lt. carcer; feogan ‘to hate’ for Lt. odisse, etc. Conversely (but perhaps more rarely), some words were originally dialect words but later became part of the common vocabulary; e.g. ModE sunset is first attested in Late Nhb. (Lindisfarne Gospels). Words may be characteristic of a certain dialect just in a specific meaning but not in all meanings. For example, soðfæst was apparently common Old English in the meaning ‘true’ Lt. verus, but Anglian in the meaning ‘just’ Lt. iustus. On modig (‘proud’ in the Winchester Vocabulary, but ‘high-spirited’ in Angl. poetry) see above. Also, a specific derivation may be characteristic of a certain dialect, whereas the basis and other derivations from the same basic word (especially prefix or suffix formations) may be characteristic of different dialects or of common Old English. For example, (ge)hreowian ‘to repent’ apparently was a general OE word, whereas behreowsian ‘to repent, to regret’ was restricted to the LWSax. texts of the Winchester Group. Similarly, ongietan ‘to understand’ was apparently a general OE word, whereas undergietan was a typical Winchester word. In the following paragraphs we list some of the better known OE dialect words. Often we also give the Lt. word(s) which were translated or glossed by the OE words in question.
11.3 West-Saxon The West-Saxon vocabulary was not a homogeneous block. A chronological distinction has to be made between Early West-Saxon and Late West-Saxon, and there were also differences (sub-dialects) within Late West-Saxon. Particularly striking is the so-called Winchester Vocabulary, see Section 12 below. As mentioned above, some words were still used in Early WSax. which were later restricted to Anglian. In Late West-Saxon, many new words and formations appeared. Examples of WSax. words are:
22 Old English: Dialects (1) nouns: cnapa, cnafa ‘child, servant’ (> ModE knave); ofermod, etc. ‘pride’ for Lt. superbia; tima > ModE time; geswinc ‘toil, effort’ for Lt. labor (Angl. gewinn); LWSax. cynehelm ‘crown’ for Lt. corona; eorðtilia ‘farmer’ for Lt. agricola (Angl. landbuend). The following loan-words are also first attested in LWSax.: lagu (from ON) (> ModE law); pryte (from OFr.) (> ModE pride). (2) adjectives: gehwæde ‘slight, small’. (3) verbs: ætbregdan ‘to take away’; behatan ‘to promise’; (ge)fægnian ‘to rejoice’; forðfaran ‘to depart, die’ (Angl. (ge)leoran); hopian > ModE hope (Angl. hyhtan); scrydan ‘to clothe, dress’; LWSax. afeormian ‘to cleanse, clean’ for Lt. purgare (common OE geclænsian > ModE cleanse).
11.4 Anglian Much of the research on OE dialect vocabulary has concentrated on Anglian: see Jordan (1906) and Wenisch (1979). Specifically Anglian words occur among all the major word classes, e.g.: (1) nouns: ambeht- ‘office’ for Lt. officium, ministerium; morðor ‘murder, homicide’ (> ModE murder); scua ‘shade, shadow’ for Lt. umbra; symbel ‘feast’; ðreat ‘crowd, group’ for Lt. turba; gewinn ‘labor’ for Lt. labor (WSax. geswinc). (2) adjectives: medmicel ‘small, little’ for Lt. parvus; soðfæst ‘just’ for Lt. iustus (3) verbs: acweðan ‘to say, tell’ for Lt. dicere; bebycgan ‘to sell’ (WSax. sellan); leoran ‘to go, pass away’ for Lt. ire, obire, praeterire, transire; frignan ‘to ask’ for Lt. interrogare; winnan ‘to labor’ for Lt. laborare. (4) adverbs, pronouns, prepositions, etc.: gen(a), geona ‘so far’ for Lt. adhuc; in (WSax. on); nænig ‘nobody, nothing’ for Lt. nemo, nihil (common OE na¯n). See also Section 6 above. The words mentioned were apparently shared by Mercian and Northumbrian. Not much is known about a specifically Mercian vocabulary (see Kastovsky 1992: 343). On specifically Northumbrian vocabulary, see the next section.
11.5 Northumbrian The three long Northumbrian texts or rather glosses from the second half of the 10th century (glosses to the Lindisfarne Gospels, Rushworth Gospels2, and the Durham Ritual; see above) which are the main witnesses for Northumbrian all represent the language of Aldred the scribe, i.e., the language of one individual. The question, therefore, is the extent to which the words confined to these texts represent contemporary Northumbrian dialect words and how far they represent Aldred’s idiolect, including his own innovations, i.e., loan translations (translation idiolect). A number of words were probably not even used by Aldred in his everyday speech. They were presumably coined by him as loan-formations (loan-translations), not in order to enrich the OE vocabulary, but rather to explain the morphological structure of complex Latin lemmata. Some of his formations may even have been meaningless without recourse to the Latin word they are meant to explain, e.g. respicere – eftbehealdan, lit. ‘again behold’. Thus there are two categories:
355
356
III Old English (1) some words seem to represent genuine Northumbrian dialect words, e.g. drysn(i)an ‘extinguish’ for Lt. extinguere and evanescere; hoga ‘prudent’ for Lt. prudens, and hogascipe ‘prudence’ for Lt. prudentia; portcwen ‘whore’ for Lt. peccatrix, meretrix; sunset > ModE sunset for Lt. occasus. (2) other words were apparently coined by Aldred as loan-formations modeled on their Latin lemmata: Aldred, for example, usually renders Latin words beginning with con-, com- by OE words beginning with efne- ‘evenly; equally’, etc., and Latin words beginning with re- by OE formations beginning with eft- ‘again’, etc. e.g. commemoratio – efnegemynd ‘remembrance’, lit. ‘evenly-remembrance’; considerare – efnesceawian ‘consider’, lit. ‘evenly-consider’; respicere – eftbehealdan and eftbeseon ‘see, behold, catch sight of’, lit. ‘again-see’, ‘again-behold’, etc.
11.6 Kentish Although the main features of Kentish phonology are well established (see Section 8 above), comparatively little is known about Kentish dialect vocabulary; it was probably never as distinctive as the Anglian and the West-Saxon vocabulary. Although Canterbury, as the starting point of the Christianization of England and as the seat of an archbishop, was one of the most important ecclesiastical centers as well as an important centre for manuscript production, in its vocabulary, it seems to have been influenced by Angl. and WSax. elements; Kent. features were apparently not stressed. At Canterbury, nobody developed the kind of standardized vocabulary that Æthelwold and his pupils did at Winchester. This is probably connected with the history of Kent: Kent was first dominated by Anglian kings, and later by WestSaxon kings. Nevertheless, a number of 11th century glosses which were copied at Christ Church, the cathedral of Canterbury, or even originated there, have some words which seem to reflect a kind of Canterbury usage, although they are basically WSax. (e.g. the Arundel Prayers, the Brussels Aldhelm Glosses, etc.) and some even belong to the Winchester Group (the Interlinear Version of the Benedictine Rule). Some examples are: twi-seht ‘quarrel’; ungecoplic ‘unsuitable’; stæfwis ‘educated’ for Lt. litteratus.
12 The Winchester Vocabulary A distinctive sub-dialect within Late West Saxon was the language of the so-called Winchester School or Winchester Group; see Gneuss (1972); Hofstetter (1987, 1988); Kastovsky (1992: 347–349); Ono (1986) (see also Kornexl, Chapter 24). The Winchester Vocabulary was probably initiated in the second half of the 10th century by bishop Æthelwold, who apparently taught it in his school at Winchester; thus it constitutes a kind of standardized school vocabulary, the product of a conscious attempt to achieve a generally accepted norm. The teacher probably said something like: for translating Lt. superbia we shall use modignes and not any other word. This vocabulary was then continued by Æthelwold’s pupils. The most important of them, due to the quality as well as the quantity of his writings, was abbot Ælfric. So far fourteen semantic groups have been established where the writers connected with or influenced by the Winchester school consistently prefer a particular word or wordfamily to their synonyms. Hofstetter classifies the words in three groups: A) typical
22 Old English: Dialects
357
Winchester words; B) neutral words, used in Winchester and elsewhere (these do not occur in all semantic groups); C) words never used in Winchester (those can be Anglian or generally West-Saxon). Here we give a slightly simplified version of his list, i.e., we give only typical representatives, not all the variant forms, see Table 22.2. Table 22.2: The Winchester Vocabulary Semantic group (1) (2) (3) (4) (5)
(6) (7) (8) (9)
(10)
(11) (12)
A) Winchester word(s)
‘strange, foreign’, ælfremed, etc. Lt. alienus, extraneus ‘martyr’, cyðere Lt. martyr, martyrium ‘to dare’, (ge)dyrstlæcan Lt. audere, praesumere ‘prepare’, (ge)gearcian Lt. (prae)parare, praebere, etc. ‘church’, (ge)laþung Lt. ecclesia (for the people, the community of believers; not for the building or the clerics) ‘virtue’, Lt. virtus miht ‘power, might, strength’, miht Lt. virtus (except sense 6) ‘fear, fright, terror’, oga Lt. terror, horror, timor, etc. ‘to direct, make (ge)rihtlæcan right, make straight, correct’, Lt. dirigere, corrigere, emendare (ge)cwysan, ‘to break, shatter, squash’, tobrytan Lt. (at)terere, confringere, (con)quassare, etc. ‘to repent’, Lt. paenitere behreowsian ‘proud, pride’, modig, Lt. superbus, superbia modignes, etc.
B) Neutral words
C) Non-Winchester words
–
fremde, afremdan, etc.
martir
þrowere
(ge)dyrstignes –
(a)þristian (ge)þristlæcung, etc. (ge)gearwian, etc.
ecclesia, geferræden
cirice, (halig) (ge)samnung
mægen –
cræft, strengu, etc. mægen, cræft, strengþ, etc. anda, egesa, gefyrhtu, etc. (ge)reccan, (ge)rehtan
broga, ege, fyrhto (ge)rihtan
–
(a)breotan, (a)brytan, etc.
– prud, prydo
hreowan, (ge)hreowian oferhygd, oferhoga, ofermede, ofermodig, etc. beag, corona, heafodbeag, etc. understandan, ongietan
(13)
‘crown, wreath’, Lt. corona
wuldorbeag
cynehelm, etc.
(14)
‘understand’, Lt. intellegere
undergietan
–
It has been pointed out (mainly by Seebold 1989a, 1989b, 1990) that the Winchester Vocabulary was apparently created from two sources: (a) Some Winchester words, especially those belonging to the more general vocabulary, were probably old and came from the dialect spoken in Winchester (according to Seebold perhaps from a Jutish substrate), (b) whereas others, especially those with theological implications, were newly formed or adapted in meaning. Latin loan-words are absent in the fourteen word-groups characteristic of the Winchester Vocabulary. To group (a), the regional
358
III Old English words, probably belong (the number given above is added in brackets): ælfremed (1), (ge)dyrstlæcan (3), gearcian (4), miht (6, 7), oga (8), (ge)rihtlæcan (9), undergietan (14). To group (b), the newly formed or semantically adapted words probably belong: cyðere (2), gelaðung (5), modig(ness) (12), wuldorbeag (13). The Winchester school was very influential; traces can be found in Canterbury and Exeter. The LWSax. vocabulary was, however, by no means identical to the Winchester Vocabulary. There were many LWSax. authors and texts which did not use the Winchester Vocabulary, among known authors notably archbishop Wulfstan and Byrhtferth of Ramsey (see the following section). It should also be noted that most of the Winchester words did not live on, but died out after the Norman Conquest. Of the fourteen words or word-groups given above, only two, namely might and moody (with change of meaning) live on. More of the neutral or non-Winchester words (B and C words) live on: church, martyr, pride (three loan-words), craft, fright, right, strength, and understand. But many of them also died out and were replaced by loan-words, e.g. behreowsian by repent, wuldorbeag by crown.
13 Wulfstan’s vocabulary That archbishop Wulfstan’s LWSax. vocabulary is quite distinct from the vocabulary of Ælfric (and thus from the Winchester Group) was shown by Jost (1950) (who, of course, did not yet know about the Winchester Group). Wulfstan uses, for example, geberan ‘create, beget’ (not -cennan); (ge)gearwian ‘prepare’ (not -gearcian); gesælig ‘happy, prosperous’ (not eadig); namian (> to name) (not hatan), etc. Seebold (1974) argues that Wulfstan represents the same subdialect of WSax. as King Alfred about a hundred years earlier. For Lt. prudens, prudentia both use wær, wærscipe, not snotor, snotornes as the Winchester group; for Lt. superbia both use ofermod or ofermettu. Wulfstan also employs a number of Scandinavian loan-words which were apparently borrowed into late Old English, e.g. lagu (> law) and derivatives such as utlaga (> outlaw); see, e.g., Pons Sanz (2007). Presumably they reflect his stay at York.
14 The question of a poetic dialect The bulk of OE poetry has come down to us in four manuscripts (Beowulf MS, Exeter Book, Junius MS, Vercelli Book; see, e.g., Ker 1957, nos. 216, 116, 334, 394) written probably by West-Saxon scribes in the late 10th century or around 1000. Most OE poems survive in just one manuscript (e.g. Beowulf, Cynewulf’s poetry, the OE biblical poems, the OE elegies, the OE riddles, etc.). Among the exceptions are Cædmon’s Hymn and Bede’s Death Song, which exist in many manuscripts and in Northumbrian as well as West-Saxon versions. (see, e.g., Krapp and Dobbie [eds.] 1931–53: Vol. VI; Robinson and Stanley [eds.] 1991). According to the traditional view most OE poems go back to earlier Anglian originals which were largely, but not entirely, West-Saxonized in the course of their transmission. In 1953, however, Kenneth Sisam (1953: 119–39) came up with the theory that there was a common and supra-dialectal OE poetic vocabulary and that, therefore,
22 Old English: Dialects vocabulary could not be used as evidence of the original dialect of OE poems, especially of poems thought to be early (see also Godden 1992: 497). This theory found many followers, including some among editors of OE poetry; perhaps one reason was that scholars were saved the trouble of trying to distinguish between Anglian and West-Saxon words in OE poems. But the research by scholars such as Schabram (1965) and Wenisch (1979) showed that Sisam’s theory is not entirely true. Poems thought to be of Angl. origin, i.e., the majority of OE poetry, including Beowulf and the poems by Cynewulf, show some typically Anglian words, whereas poems thought to be of WSax. origin, e.g. the Metres of Boethius (traditionally thought to have been composed by King Alfred shortly before 900), Genesis B, and the Battle of Maldon (composed shortly after 991) use some typically WSax. words. Thus Beowulf and Cynewulf’s Juliana use Angl. oferhygd- for ‘pride’, whereas the WSax. poems use ofermettu or ofermod for ‘pride’. This does not exclude the possibility that WSax. poets were influenced by the model of Angl. poetry (see the following paragraph), but it proves that WSax. poets also used specifically WSax. words. (See also Fulk, Chapter 25.) Certainly there was also a specifically poetic vocabulary, i.e., many OE words were apparently only used in poetry and never in prose, e.g. beadu ‘battle’, and compounds formed with it, such as beadu-rinc ‘warrior’, beadu-lac ‘war-play’, beadu-rof ‘strong in battle’ or synonyms for ‘man, warrior’, such as beorn, guma, hæleþ, rinc, secg (see, e.g. Godden 1992: 498; Kastovsky 1992: 351–352). Conversely, some prose words were not used in poetry; this does not seem to have dialectal significance, however.
15 Further development After the Norman Conquest of 1066, Old English gradually changed into Middle English, although OE MSS continued to be copied for c.150 years, i.e., until c.1215. The West-Saxon standard collapsed and the dialects became prominent once again; for the ME dialects (see Williamson, Chapter 31). A new standard only began to emerge in the second half of the 14th century, based on the language of London (itself having changed from WSax. to East Midland). Many OE words died out and French loan-words came in. When teaching Old English today, the LWSax. standard (as exemplified by Ælfric and Wulfstan) is usually taken as the basis; this is justified by the quantity as well as the quality and relative regularity of the material. But for the further development of English, especially its phonology (pronunciation), Anglian was much more important. To give just a few examples: In eald, ceald etc. WSax. shows the effects of breaking (see Section 8.3 above); this would have yielded *eeld, *cheeld, ModE /iːld/, /tʃiːld/; the Angl. forms, however, developed regularly into the ModE forms: Angl. cald > ca¯ld > co¯˛ ld > cold /kəʊld/; ald > a¯ld > o¯˛ ld > old /əʊld/. WSax. had ciese (due to palatal diphthongization, see Section 8.4 above), whereas Angl. had cese > ModE cheese /tʃiːz/. WSax. has hieran, whereas Angl. had heran > ModE hear /hiːə/.
16 References Bately, Janet. 2009. Did King Alfred actually translate anything? The integrity of the Alfredian canon revisited. Medium Aevum 78: 189–215.
359
360
III Old English Baugh, Albert C. and Thomas Cable. 2002. A History of the English Language. 5th edn. London: Routledge. Becker, Alfred. 1973. Franks Casket: Zu den Bildern und Inschriften des Runenka¨stchens von Auzon. Regensburg: Hans Carl. Brunner, Karl. 1965. Altenglische Grammatik nach der angelsa¨chsischen Grammatik von Eduard Sievers neubearbeitet. 3rd edn. Tu¨bingen: Niemeyer. Campbell, Alistair. 1959. Old English Grammar. Oxford: Oxford University Press. Colgrave, Bertram and R. A. B. Mynors (eds.). 1969. Bede’s Ecclesiastical History of the English People. Oxford: Clarendon. Frantzen, Allan J. 1986. King Alfred. Boston: Twayne. Gneuss, Helmut. 1955. Lehnbildungen und Lehnbedeutungen im Altenglischen. Berlin: Schmidt. Gneuss, Helmut. 1972. The origin of Standard Old English and Æthelwold’s School at Winchester. Anglo-Saxon England 1: 63–83. Gneuss, Helmut. 2001. Handlist of Anglo-Saxon Manuscripts. Tempe: Arizona Center for Medieval and Renaissance Studies. Godden, Malcolm. 1992. Literary language. In: Hogg (ed.), 490–535. Godden, Malcolm. 2007. Did Alfred write anything? Medium Ævum 76: 1–23. Hines, John. 1991. Some observations on the runic inscriptions of early Anglo-Saxon England. In: Alfred Bammesberger (ed.), Old English Runes and their Continental Background, 61–83. Heidelberg: Winter. Hofstetter, Walter. 1987. Winchester und der spa¨taltenglische Sprachgebrauch. Munich: Fink. Hofstetter, Walter. 1988. Winchester and the standardization of Old English vocabulary. AngloSaxon England 17: 139–161. Hogg, Richard M. 1992a. A Grammar of Old English. Vol. I Phonology. Oxford: Blackwell. Hogg, Richard M. (ed.). 1992b. The Cambridge History of the English Language. Vol. I. The Beginnings to 1066. Cambridge: Cambridge University Press. Hogg, Richard M. and David Denison. 2006. A History of the English Language. Cambridge: Cambridge University Press. Jordan, Richard. 1906. Eigentu¨mlichkeiten des anglischen Wortschatzes. Heidelberg: Winter. Jost, Karl. 1950. Wulfstanstudien. Bern: Francke. Kalbhen, Ursula. 2003. Kentische Glossen und Kentischer Dialekt im Altenglischen. Frankfurt am Main: Lang. Kastovsky, Dieter. 1992. Semantics and vocabulary. In: Hogg (ed.), 290–407. Kelly, S. E. 1999. The Electronic Sawyer, Online Catalogue of Anglo-Saxon Charters. http://www. trin.cam.ac.uk/chartwww/eSawyer.99/eSawyer2.html Ker, Neil R. 1957. Catalogue of Manuscripts Containing Anglo-Saxon. Oxford: Oxford University Press. Kitson, Peter. 1995. The nature of Old English dialect distributions, mainly as exhibited in charter boundaries. Part I. Vocabulary. In: Jacek Fisiak (ed.), Medieval Dialectology, 43–135. Berlin/ New York: Mouton de Gruyter. Korhammer, Michael. 1980. Altenglische Dialekte und der Heliand. Anglia 98: 85–116. Krapp, George Philip and Elliott van Kirk Dobbie (eds.). 1931–1953. The Anglo-Saxon Poetic Records (ASPR), 6 vols. New York: Columbia University Press. Lapidge, Michael, John Blair, Simon Keynes, and Donald Scragg (eds.). 1999. The Blackwell Encyclopaedia of Anglo-Saxon England (BEASE). Oxford: Blackwell. von Lindheim, Bogislav. 1972. Das Suffix -bære im Altenglischen. Archiv fu¨r das Studium der neueren Sprachen 208: 310–320. Luick, Karl. 1921. Historische Grammatik der englischen Sprache. Band 1: 1. Abteilung. Stuttgart: Tauchnitz. Nielsen, Hans F. 1985. Old English and the Continental Germanic Languages: A Survey of Morphological and Phonological Interrelations. 2nd edn. Innsbruck: Institut fu¨r Sprachwissenschaft [1st edn. 1981].
22 Old English: Dialects
361
Nielsen, Hans F. 1991. The Straubing Heliand-fragment and the Old English dialects. In: P. Sture Ureland and George Broderick (eds.), Language Contact in the British Isles, 243–273. Tu¨bingen: Niemeyer. Ono, Shigeru. 1986. Undergytan as a ‘Winchester’ word. In: Dieter Kastovsky and Aleksander Szwedek (eds.), Linguistics across Historical and Geographical Boundaries: In Honour of Jacek Fisiak on the Occasion of his Fiftieth Birthday. Vol. I, 569–577. Berlin/New York: Mouton de Gruyter. Park, Young-Bae and Gaby Waxenberger. 2011. Old English runes and runic inscriptions: The state of the art. Poetica (Special issue: From Runes to Caxton: Aspects of Medieval English Language and Literature, ed. by Hans Sauer and Michiko Ogura) 75: 1–28. Pons-Sanz, Sara. 2007. Norse-Derived Vocabulary in Late Old English Texts. Wulfstan’s Works: A Case Study. Odense: University Press of Southern Denmark. Robinson, F. C. and E. G. Stanley (eds.). 1991. Old English Verse Texts From Many Sources: A Comprehensive Collection. Copenhagen: Rosenkilde and Bagger. ¨ bersetzungen, zusammen mit Sauer, Hans. 1978. Theodulfi Capitula in England: Die altenglischen U dem lateinischen Text herausgegeben. Munich: Fink. Sauer, Hans. 1992. Old English word geography: Some problems and results. In: Wilhelm G. Busse (ed.), Anglistentag 1991 Du¨sseldorf: Proceedings, 307–326. Tu¨bingen: Niemeyer. Sawyer, Peter (ed.). 1968. Anglo-Saxon Charters. An Annotated List and Bibliography. London: Royal Historical Society. Schabram, Hans. 1965. Superbia: Studien zum altenglischen Wortschatz. Munich: Fink. Schabram, Hans. 1970. Bemerkungen zu den altenglischen Nomina agentis auf -estre und -icge. Anglia 88: 94–98. Seebold, Elmar. 1974. Die altenglischen Entsprechungen von sapiens und prudens: Eine Untersuchung u¨ber die mundartliche Gliederung der altenglischen Literatur. Anglia 92: 291–333. Seebold, Elmar. 1989a. Winchester und Canterbury: Zum spa¨taltenglischen Sprachgebrauch. Anglia 107: 52–60. Seebold, Elmar. 1989b. Die altenglischen Verben auf -læcan. In: Karin Heller, Oswald Panagl, and Johann Tischler (eds.), Indogermanica Europaea: Festschrift fu¨r Wolfgang Meid, 333–357. Graz: Institut fu¨r Sprachwissenschaft. Seebold, Elmar. 1990. Was ist ju¨tisch? Was ist kentisch? In: Alfred Bammesberger and Alfred Wollmann (eds.), Britain 400–600: Language and History, 335–352. Heidelberg: Winter. Sievers-Brunner. 1965; see Brunner, Karl. 1965. Sievers, Eduard. 1882. Angelsa¨chsische Grammatik. Halle: Niemeyer. Sisam, Kenneth. 1953. Studies in the History of Old English Literature. Oxford: Oxford University Press. Sweet, Henry . 1887. A Second Anglo-Saxon Reader: Archaic and Dialectal. Oxford: Clarendon Press. Toon, Thomas E. 1992. Old English dialects. In: Hogg (ed.), 409–451. Vleeskruyer, R. 1953. The Life of St. Chad: An Old English Homily. Amsterdam: North-Holland. Waxenberger, Gaby. 2010. Towards a Phonology of Old English Runic Inscriptions and an Analysis of the Graphemes. Unpublished Habilitationsschrift, LMU Munich. (Publication in preparation) Wenisch, Franz. 1979. Spezifisch anglisches Wortgut in den nordhumbrischen Interlinearglossierungen des Lukasevangeliums. Heidelberg: Winter.
Hans Sauer and Gaby Waxenberger, Munich (Germany)
362
III Old English
23 Old English: Language contact 1 2 3 4 5
Introduction Popular/non-religious words Language contact as a result of conversion to Christianity Summary References
Abstract Latin loanwords came into Old English in two different ways: either through direct contact of the Anglo-Saxons (or of those Germanic tribes who later were to become the AngloSaxons) with the Romans or other Latin-speaking populations (up to about 600 CE) or through the conversion of the Anglo-Saxons to Christianity (from 600 onwards to the end of the Anglo-Saxon period). Loanwords that came into Old English during the first period fall into three groups: (1) words that are common to both Old English and Old High German, and identical phonologically and semantically; (2) words that are common to both Old English and Old High German, phonologically identical but with different meanings; and (3) words existing in Old English and having no parallels in Old High German. When the Anglo-Saxons adopted Christianity, they dealt with the new Christian vocabulary in three different ways: (1) they used existing words, but charged them with new meaning; (2) they created calques (loan translations); and (3) they adopted loanwords.
1 Introduction Three major languages exerted an influence on Old English: Celtic, Latin, and Old Norse. This chapter will focus on the influence Latin had on Old English (see Dance, Volume 2, Chapter 110). A search for any such influence needs to take into account a period of approximately eight hundred years, starting at the time when Germanic tribes on the Continent first came into contact with Romans, continuing into the period when Germanic mercenaries served in the Roman armies, and ending with the period in which the Anglo-Saxons converted to and practiced Christianity. The influence Latin exerted on Old English is almost entirely limited to vocabulary. In some Old English (OE) translations of Latin texts Anglo-Saxons use the “dative absolute” in imitation of the Latin “ablative absolute” (Mitchell and Robinson 1992: 106), but this type of syntactic construction is rarely, if ever, found in texts that are not translations. In the absence of any significant and lasting syntactic and/ or morphological influences of Latin on Old English, this chapter shall concentrate on lexical influence. Latin words came into English through two distinct channels: either through direct contact with the Romans or with Latin-speaking Gauls and Britons (up to about 600 CE), or through the book-Latin of the Christian clergy (up to the end of the Anglo-Saxon period, i.e. 1066, or, more generally, approximately 1100). It would, of course, be wrong to speak of “Old English” at the beginning of this period; nonetheless when the tribes who later were to become the Anglo-Saxons (whose language is Old Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 362–373
23 Old English: Language contact English) were still on the Continent, they adopted Latin loanwords and retained them when they settled on the British Isles. At this stage, no distinction can as yet be made between Old English speakers and those who spoke the other West Germanic (WGrmc.) languages, and hence I shall use the more general term West Germanic for this early period. According to Bede, the tribes who were later to become the Anglo-Saxons migrated to the British Isles in 449 in response to an invitation from the British king Vortigern who was harassed by the Picts and who could no longer rely on Roman troops who had protected southern Britain against any attackers up to about 410 but had withdrawn from the British Isles shortly thereafter (Bede, History [Sherley-Price, trans. 1993]: 50, 51–53, 55–57). If we are to believe Bede, these Germanic-speaking tribes from what is now northern Germany, Holland, and possibly southern Denmark came to the island after the Romans had left. The predominant language in Britain would therefore have been Celtic and to a limited extent Latin, used by the Celtic clergy and possibly by some Celtic nobility who wished to retain the Roman ways. Archaeological evidence, however, shows that at least some of the continental Saxons had arrived in Britain as early as the second half of the 4th century (Capelle 1990: 11), though it is not entirely clear whether they arrived as “laeti”, i.e. relocated prisoners-of-war who were given land to settle (Myres 1986: 74–103), or as “foederati”, i.e. mercenaries in the Roman army (Capelle 1990: 11; also Capelle 1998: 59–60). In either case, however, they would have had at least some direct exposure to the Latin tongue. Most recent scholarship argues that because of the similarities of the Latin spoken in Gaul and in Britain in the periods from about 300 to 600 it is next to impossible to determine whether a Latin loanword was adopted on the Continent or in Britain (Gneuss 1993; Wollmann 1993). This argument seems convincing since the continental Saxons were “laeti” and “foederati” of the Romans as early as the second half of the 3rd century, though not necessarily in Britain (Capelle 1998: 59), and would have retained any loanwords from Latin when they emigrated to Britain. Nor were the words phonologically so different that one could distinguish between them. Neither the Saxon spoken on the Continent nor the Saxon spoken in Britain was ever written down in any lengthy documents, and hence linguists have little direct evidence for a Latin loanword’s first appearance in the WGrmc. languages. The generally accepted test for determining whether a Latin loanword in Old English has come into the language at this early period is an examination of whether the same loanword exists in the same lexical meaning in other WGrmc. languages, and in approximately the same phonological form, though, of course, sound laws in both Old English and the other WGrmc. languages could intervene (Nielsen 1998: 147). Below I shall limit myself to Old High German (OHG) words for comparison with Old English words since a Latin loanword common to both Old High German and Old English suggests that the word came into West Germanic when Old High German and Old English had not developed too far from each other. Writing arrived in England with the conversion of the Anglo-Saxons to Christianity in 597 and was used for the vernacular as early as the beginning of the 7th century (Greenfield and Calder 1986: 107). From that time on, the evidence for Latin words coming into Old English is more secure, as long as one remembers that the first spoken occurrence of a Latin loanword in Old English may predate the first written occurrence by as much as a decade, a century, or even several centuries. Conversely, a written Latin loanword is no guarantee for that word also being used in the spoken language. Latin
363
364
III Old English loanwords first appear in glossaries, and later in the works of Alfred, whose educational reform stipulated that translations be made of certain Latin works, books which he considered “most needful for all men to know” (Alfred, Gregory’s Pastoral Care [Sweet, ed. 1871]: 6). Towards the end of the Anglo-Saxon period, the writer Ælfric, who flourished around the year 1000 (Zupitza [ed.] 2000 [1880]), wrote a grammar and a large number of Old English homilies and saints’ lives, and since these too are based on Latin models, they provided an opportunity for Latin loanwords being introduced into Old English. There is, then, a basic distinction in the types of words being introduced into Old English: up to about 600, the words would be mostly “popular”, i.e. non-religious words denoting everyday objects and concepts, while the words being introduced after 600 would be mostly “learned”, i.e. religious words and those denoting objects and concepts connected with book learning (Gneuss 1993: 113, referring to Pogatscher 1888). On the phonological differences between Latin loanwords adopted into Old English before and after c.600, see Campbell (1959: 200–219).
2 Popular/non-religious words 2.1 Latin loanwords common to both Old English and Old High German, and identical phonologically and semantically Latin loanwords common to Old English and Old High German and retaining the same meaning in both languages can with reasonable certainty be assumed to have been borrowed by the WGrmc. speakers directly from the Romans, provided the phonology is near-identical or the phonological differences can be accounted for through later sound shifts. Thus, even though e.g. the Old English and OHG words ceapman and koufman, both meaning ‘merchant’, do not immediately appear to be related, both in fact derive from the Lt. loanword caupo, to which both languages added -man. WGrmc. /au/ develops to /æa/ (spelled ) in Old English (Campbell 1959: 203), and the initial , pronounced /k/ in Latin, becomes palatalized before the front vowel, giving /tʃ/ (Campbell 1959: 65). Hence Lt. caupo changes to OE ceap, pronounced /tʃæap/, which becomes the ancestor of the chap part in Modern English (ModE) chapman. In the OHG version of the Lt. loanword, the initial /k/ is retained, as is the /au/, but the second Germanic consonant shift, taking place some time in the 6th to 8th centuries moves the /p/ to an /f/ (on the second Germanic consonant shift, see Bach 1965: 101–112; Lockwood 1976: 51–56). As this example shows, Lt. loanwords that come into the WGrmc. languages at an early period are fully subject to all the same sound changes that native words undergo. The Old English text Beowulf (Fulk et al. [eds.] 2008) has about ten such Latin loanwords, parallels of which also exist in Old High German and can therefore be accepted as words which the WGrmc. people borrowed directly from the Romans in the period prior to 450 CE. Beowulf seems appropriate for a study of such loanwords because it is a secular text with no direct religious concerns, and hence would seem more likely to reflect the vocabulary used and understood by the general populace rather than the more learned vocabulary of the clergy. Table 23.1 provides the Latin form which the WGrmc. speakers borrowed, the OE and OHG versions, as well as the ModE and Modern German (ModGer.) variants, if they still exist. If Modern English now uses a different word, it is provided in square brackets.
23 Old English: Language contact
365
Table 23.1: Latin Borrowings in Beowulf Latin
Old English
OHG
ModE
ModGer.
ancora campus caupo discus draco gemma milia strata vallum vinum
ancor camp ceap disc draca gim mil stræt weall win
anker champf kouf tisc trahho gimma mila straza wal win
anchor [fight] chap dish drake [gem] mile street wall wine
Anker Kampf Kauf Tisch Drache Gemme Meile Straße Wall Wein
A few brief comments on some of some of the words in Table 23.1: – On ancora, see Frank (2001: 7–27). – The Lt. campus actually means ‘field, battlefield’ rather than ‘fight’, but both OE camp and OHG champf through metonymic shift refer to the activity on the battlefield rather than to the place itself (except in place names, where the -camp/comb part often refers to ‘field’; see Gelling 1977: 5–8). – The ModE gem clearly derives from Lt. gemma, but it has come into Modern English not via Old English but via French (Fr.). The OE word was replaced by the French word after the Norman invasion. – Lt. milia is a numeral and means ‘thousand’; a thousand double steps make a mile, and hence the simple numeral became a measure of length. – The OHG straza keeps the Lt. /a/, and the original /t/ becomes an /s/ in the second Germanic consonant shift. In Old English, through Anglo-Frisian fronting the /a/ of strata becomes an /æ/ (Campbell 1959: 52, 203), but retains the /t/. – Similarly, Lt. vallum shows first Anglo-Frisian fronting to */wæll/ and then, through breaking, shifts to weall (Campbell 1959: 55–56). Most of these words have to do with the areas of trade (ancora, caupo, discus, gemma, vinum) and war (campus, draco [on the standards of the Roman army], and vallum), with the streets (strata), that were measured in miles (milia), serving for both mercantile and military purposes. Beowulf has a few more Latin loanwords that happen to be identical (taking into account sound changes) in form and meaning in both Old High German and Old English. These words are OE deofol (‘devil’) and OE gigant (‘giant’), but since both are strongly associated with Christianity, they will be discussed further below (see Section 3).
2.2 Latin loanwords common to both Old English and OHG, phonologically identical but with different meanings Beowulf also contains two Latin loanwords for which the Old High German has a different meaning. These words are scrifan (derived from Lt. scribere, ModE shrive), and segn (derived from Lt. signum, ModE sign, which in turn derives from Fr. signe and not
366
III Old English from OE segn). OE scrifan does exist in OHG as scriban (for other words that have OHG and Old English see, e.g. OHG geban and OE giefan [ModE to give] or OHG lioben and OE lufian [ModE to love], neither of which are Lt. loanwords), but with a different meaning: whereas the OE word originally means ‘to prescribe, ordain, assign, impose’ and in the later OE period ‘to hear confession, to receive absolution’, the OHG word means ‘to write’ and has retained that meaning into Modern German. The OE segn also exists in Old High German as segan, but whereas the Old English translates to ‘ensign, banner’, the Old High German translates to ‘blessing [derived from the “sign of the cross”]’. Why do the words scrifan/scriban and segn/segan have such different meanings in the two languages? Neither of the OE words, at least in the forms and meanings in which they appear in Beowulf, has any religious connotations, and thus does not owe anything to the arrival of Christianity in Anglo-Saxon England. From historical records we know that the Romans employed Germanic speakers, among them Saxons, as mercenaries and employed them in the provinces Belgica II and Germania II (Capelle 1998: 63), and it would be surprising if these mercenaries, who had daily contact with their Roman overlords, did not pick up a few Latin loanwords. Unfortunately, these mercenaries did not write down their language, and hence it is next to impossible to determine the extent to which they let Latin slip into their own every-day speech. The words they would be most likely to take over from Latin presumably would be those denoting military equipment, ranks and units, and quite possibly commands, but if that was the case, most of these words did not survive until today. Modern English, for instance, uses sword, derived from OE sweord, and not a variant of the Lt. gladius. Similarly, it uses shield and helmet from OE scield and helm rather than from Lt. scutum and cassis. And where Modern English has Latin-derived words for military equipment or units, such as lance (from Lt. lancea), legion (from Lt. legio), or cavalry (ultimately from Lt. caballus ‘horse’), it quickly becomes apparent that these words are much later borrowings, all three from French during or after the Middle English period. Some or most of the Latin words borrowed by the early Saxon mercenaries may, of course, have become defunct over the centuries and no longer exist in Modern English, but even an examination of the OE lexicon does not yield many such terms. Two possible exceptions may be the above-mentioned OE words segn and scrifan. One possible explanation for the difference in the meanings of the identical loanword in the two languages consists in the OHG speakers taking their segan from the Latin clergy, and the OE speakers retaining the Latin loanword in the meaning which they had received as mercenaries from their Latin military superiors. Similarly with the word scrifan: the Germanic mercenaries would see Roman writing primarily as their officers’ means of prescribing, ordaining, assigning, or imposing, and hence would use the word in this secondary meaning rather than, as the OHG speakers, in its primary meaning of ‘writing’. With these two words, the OE speakers seem to have retained a meaning which their ancestors first encountered when they served as mercenaries in the Roman armies.
2.3 Latin loanwords existing in Old English only with no parallel in Old High German Beowulf also contains several Latin loanwords that do not exist in Old High German. These are candel, non, and ceaster, of which the first two will be discussed below (see
23 Old English: Language contact Section 3). Ceaster survives in Modern English primarily in place names such as Doncaster, Leicester, and Winchester, all of which name former Roman fortifications. The -caster, -cester, and -chester part of these names derives from Lt. castra. In Old English the word ceaster also existed independently, i.e. not just as a component in a place name, in the meaning of ‘castle, fort, town’. The Anglo-Saxons may have taken the place names over from the Celtic inhabitants, in which case the Lt. loanword comes into Old English through Celtic. The fact, however, that Old English also has ceaster as an independent word opens up the possibility that the Saxons, either on the Continent or already in Britain, took over this word directly from the Romans. In some ways, the WGrmc. speakers had no need for a Latin loanword since they possessed a native word for fortification, namely OHG burug and OE burh, the latter of which develops into -borough, -burgh, or -bury in Modern English, and into -burg in German (e.g. Augsburg, Regensburg, both places that had Roman castra). And yet they did so, possibly to distinguish between the Germanic and the Roman types of fortification (the Germanic people using wooden palisades and the Romans using stone walls). There can be little doubt that ceaster is an early borrowing. The Lt. castra experiences Anglo-Frisian fronting of the /a/ > /æ/, palatalization of the /k/ to /tʃ/, and diphthongization of the /æ/ to /æa/. Despite the fact that it is an early borrowing, however, castra is not a word that was borrowed by both the continental Germans and the (Anglo)Saxons (Nielsen 1998: 158). The Romans did have castra, i.e. fortifications, on German soil just as they did in Britain, but Old High German took as its loanword not a form of castra, but a form of castellum, e.g. in the place names Kassel and Bernkastel, and a derivative of castellum later came into English from the French in the word castle. The difference between the continental castellum and the insular castra allows for the possibility that the (Anglo)-Saxons adopted the word castra only after they had arrived in Britain, though whether they learned it from Romans or Latin-speaking Celts is impossible to decide (Jackson 1953: 252).
3 Language contact as a result of conversion to Christianity Saxon mercenaries in the Roman army, either on the Continent or in Britain, would undoubtedly come into contact with fellow soldiers who were Christian. While there is no evidence of wide-spread conversion of the Saxons at this period, at least some mercenaries inevitably would become aware of the Christian clergy, its rituals, some of its teachings, and some of the implements used during the mass, and hence would acquire some vocabulary denoting these concepts and things. The full extent of this language contact, however, cannot be determined, primarily because of lack of any written sources. The basic rule enunciated above for non-religious terms can be invoked here as well, namely that a loanword seems to have been taken over prior to about 500 if it exists in both Old English and Old High German, but that rule needs to be applied with even greater caution for religious terms for two reasons: one, the missionaries to the Anglo-Saxons may well have employed terms that were already accepted by Christian Germanic speakers on the Continent (see Bede, History [Sherley-Price, trans. 1993]: 69 on “interpreters from among the Franks” being present at Augustine’s mission), and hence a common term for Old English and Old High German does not necessarily mean that those two groups accepted the loanword at the same time; and two, starting in the late 7th century, AngloSaxons went over to the Continent as missionaries, and this contact may well have
367
368
III Old English brought about a harmonizing of the religious language of the OE and OHG speakers on the Continent, and one that would also have an influence on Old English. The traditional date for the start of the conversion of the Anglo-Saxons to Christianity is 597, the year in which Pope Gregory the Great sent Augustine to Canterbury (Bede, History [Sherley-Price, trans. 1993]: 68–70). The actual conversion of the Anglo-Saxons is somewhat more complex, however. Æthelbert of Kent had married the Merovingian princess Bertha, a Christian who had brought a Christian bishop and possibly priests with her (Bede, History [Sherley-Price, trans. 1993]: 69). It would therefore be safe to assume that some talks about religion took place between Anglo-Saxons and Frankish clergymen before 597. Moreover, the Celtic Britons, the Welsh, and the Picts in the North all were Christians, and even though Bede asserts that the Britons refused to convert the Anglo-Saxons (Bede, History [Sherley-Price, trans. 1993]: 66), some Picts and some Welsh may have done so before the “official” date. In any case, some Christian vocabulary is likely to have come into Old English prior to 597, so that it is probably best to speak of the Christian Latin influence on Old English as beginning in the 6th century. This influence is anything but uniform. The language of Christianity in the West was Latin, and when the missionaries wished to communicate the religious concepts to the laity, they had the choice of a) using words already existing in Old English, though charging these words with a new meaning; b) creating “calques” (loan translations), i.e. translating Latin words literally into Old English; or c) adopting the Latin word as a loanword.
3.1 Using words already existing in Old English, though charging these words with a new meaning A few examples may suffice. The term deus ‘god’ was rendered with a Germanic word, though in the process the concept of ‘god’ changed considerably, denoting a monotheistic rather than a polytheistic god. Similarly, the Latin dominus ‘lord’, was rendered as drihten or frea, and though in one of its meanings it still referred to a secular lord, once applied to the divinity, it underwent a considerable change to the divine ‘Lord’. Other words falling in this category are the translations for Lt. gehenna ~ OE hell/ModE hell (from the Germanic goddess of the underworld, Hel), Pascha ~ OE Easter/ModE Easter (from the Germanic goddess of spring, Eostre), sanctus ~ OE halig/ModE holy (derived from hal ‘whole, healthy’), crux ~ OE rod/ModE rood (rod was the word for “cross” used during most of the OE period), eucharistia ~ OE husl/ModE housel (husl originally signified ‘sacrifice’), and peccatum ~ OE syn/ModE sin (syn originally meant ‘injury, mischief, feud, guilt, crime’). It is hardly surprising that the clergy sought to find equivalents for the Latin terms in the language of the Anglo-Saxons, but the choice of two Germanic goddesses for Christian concepts is unexpected. It should be mentioned here that several of the above terms are almost identical in Old High German, such as god = got, Easter = ostarun, hell = hella, syn = sunta, and halig = heilag. This fact might lead to the conclusion that these words were already fixed in their new meaning on the Continent and that the Anglo-Saxons brought these words from there (i.e. prior to about 500). Although this
23 Old English: Language contact possibility cannot be entirely excluded, the question arises why pagan Anglo-Saxons would have brought these Christian words with them to England where they would have had no need or even opportunity to use them. Considering the continued cultural contacts between England and the Continent throughout the 6th and later centuries, it seems more likely that the missionaries would not invent new translations but introduce those that were already in use among Germanic speakers, such as god, Easter, Hell, and sin. For at least two of the above words we know that the direction of the influence was from England to the Continent: Anglo-Saxon missionaries in the 8th century took the OE halig to the Continent, where it displaced the OHG wih (cf. words such as Weihnacht [‘Christmas’], literally ‘Holy night’; Weihrauch [‘incense’], literally ‘holy smoke’; or the place name Weihenstephan ‘St. Stephen’; see Bach 1965: 158; Braune 1918: 400; Bach 1965: 158 also mentions easter as a word brought by Anglo-Saxon missionaries to Germany). Continued cultural exchanges into the 8th century between the Continent and England are a more likely explanation for identical words translating Latin Christian concepts than the assumption that the pagan Saxons somehow brought these Christian words to England from the Continent.
3.2 Creating “calques” (loan translations) Prior to the coming of Christianity to England, the Anglo-Saxons had no need to find equivalents for such Lt. words as trinitas (‘Trinity’), evangelium (‘gospel’), discipulus (‘disciple’), omnipotens (‘almighty’), dies judicii (‘Dooms Day’), or patriarchus (‘patriarch’). There were no convenient native words that could have expressed these concepts. The Anglo-Saxons could have taken these Latin words over in their Latin forms as loanwords, but they did not do so. Instead, they chose a middle way, i.e. they translated the components of the Latin words literally into Old English and thus created words (or lexical units) that had never existed in Old English before. These loan translations are also called “calques”. Most often these calques literally translate the two concepts of the Latin word. Thus the OE godspel, from god ‘good’ and spel ‘narrative, history, tale, fable, message, news’, literally translates the two component parts of the Lt. (originally Gk.) evangelium, the first of which is eu-, meaning ‘good’, and the second angelium, meaning ‘message’. Similarly, the Lt. omnipotens, consisting of omni- ‘all’ and potens ‘powerful, mighty’ is translated into Old English as ælmihtig, combining OE æl- ‘all’ and mihtig ‘mighty’. In these examples, nouns plus adjectives or adjectives plus adjectives are combined, but occasionally Old English can also combine an adjective and a suffix, especially when the Latin has the same pattern. This happens for the OE equivalent to the Lt. trinitas, which is ðrines, consisting of ðrie ‘three’ and the suffix -nes, which Modern English still has in words such as dark-ness, numbness, red-ness. The Lt. discipulus, from discere ‘to learn’, is rendered into Old English as leorning cniht, literally a ‘learning boy’, which translates the suffix -pulus etymologically correctly with ‘boy’ (the suffix derives from the root word for puer ‘boy’; Lewis and Short, s.v. discipulus). Not all translations are totally accurate. The Lt. patriarchus, which in turn derives from Greek, has as its components pater ‘father’ and arche ‘beginning, leadership’ and the OE rendering of this word heah fæder, literally ‘high father’ seems to confuse Gk. arche with Gk. akros ‘high’. Most calques, though, are correctly translated, and their staying power is attested by the fact that words such as Almighty, gospel, and Doomsday are still retained in Modern English.
369
370
III Old English
3.3 Adopting loanwords 3.3.1 Religious loanwords common to Old English and Old High German Old English and Old High German have some religious Latin loanwords in common, and they agree semantically as well as phonologically (taking into account sound changes). Two of these words we have already encountered in Beowulf: OE deofol and gigant (OHG tiufal and gigant, ModE devil and giant, whereby the ModE term derives not from Old English, but from Fr. geant), and there are more, though not in Beowulf: OE cirice/OHG kiricha (ModE church), OE biscop/OHG biscof (ModE bishop), OE preost/OHG prestar, and OE engel/OHG engil (ModE angel). All of these words are originally Greek, but with one exception were taken over as loanwords in Latin in the forms diabolus, gigas (with oblique forms in gigant-), episcopus, presbyter and angelus, and came into the Germanic languages from Latin (with presbyter being shortened to prestar in Old High German and further shortened to preost in Old English). The one exception is OE cirice/OHG kiricha, which derives from Gk. kyriako´n ‘[house] belonging to the Lord’, which was not borrowed by Latin speakers (their word is ecclesia, itself derived from Greek). Let us deal with cirice/kiricha first: the word appears to have been used in the Lyon area in Southern Gaul, and spread from there first to Latin-speaking Trier in the 4th century and from Trier into the Germanic speaking world as a loanword (Gneuss 1993: 121; relying on Masser 1966: 17–25, Mu¨ller and Frings 1966–1968: 228–232, and Scha¨ferdiek 1984: 46–50). Churches would no doubt have impressed the pagan Germanic speakers on account of their size and architectural grandeur, and hence it is not difficult to accept the likelihood that the Germanic speakers would have taken over the term denoting the impressive building (though it is far from clear why they did not adopt the word more commonly used in Romance-speaking countries, namely ecclesia). Similarly, through their robes and retinues bishops and priests would be distinguished from both the common man and the nobility, and the pagans would easily have adopted the word they heard the Romans use to speak about them. Why, however, the pagans would have adopted words denoting religious concepts such as devil and angel is harder to explain. For these one has to resort to the same reasons as mentioned above for Hell, Easter, sin, and God: continued cultural exchanges between the pagan Anglo-Saxons and the Continent, a desire by the Christian missionaries to create a uniform vocabulary for the newly converted Germanic pagans, and the Anglo-Saxon missionary activity on the Continent all combined to create some verbal overlap between Old English and Old High German.
3.3.2 Religious loanwords in Old English but not in Old High German Two Latin loanwords in Beowulf, candel and non, both taken from the religious sphere, have no equivalence in Old High German, and thus seem to have been accepted into Old English at a time when the speakers of these languages had less contact, i.e. in the 9th century and later when Anglo-Saxon missionaries no longer went to the Continent. Candel derives from Lt. candela; the lack of i-mutation (*candil > *cendel as in *angil > engel) suggests that this is indeed a later borrowing (Campbell 1959: 205). The OHG equivalent to candel is cherza. Non is a word from the monastic sphere, denoting the ninth hour (Lt. nona [hora]), i.e. 3 pm, which gradually, though after the Anglo-Saxon period, came to
23 Old English: Language contact shift to noon, i.e. 12 o’clock. The OHG equivalent is mittilatag ‘midday’. These two loanwords differ from the loanwords listed in the following category because they are not “learned” words. It seems apparent that they were not borrowed on the Continent nor were they harmonized with Christian words from the Continent, but once taken over into Old English, they quickly became “everyday” words.
3.3.3 Learned or religious loanwords The largest influx of Latin loanwords into Old English occurred in the later part of the Anglo-Saxon period, primarily in the translations made directly or commissioned by Alfred (in the last two decades of the 9th century), and then again in the grammar, homilies and saints’ lives written by Ælfric (around the year 1000). Both Alfred’s and Ælfric’s translations were based on Latin originals and the difficulty in always finding a native word to express the Mediterranean concept led to their adopting Latin loanwords. In the translation of Orosius’s Historia adversus paganos, which appears to have been commissioned by Alfred, we encounter, among others, the following loanwords: consul, legie, senatus, talente, triumphe, and tictator (‘consul’, ‘legion’, ‘senate’, ‘talent’, ‘triumph’, and ‘dictator’) (Orosius [Bately, ed. 1980]). All of these terms indicate Roman institutions or monetary units and are untranslatable. Modern English still has these same terms, although, as the differences between OE legie and ModE legion, OE tictator and ModE dictator show, not directly from Old English, but from French and Latin respectively. At times, the Latin loanwords seem to ignore cultural differences, when the translator, for instance, renders the word for ‘pagan priests’, namely pontifices, with the OE loanword bisceapas, or when he refers to a Vestal Virgin with the OE loanword nunne (from late Lt. nonna ‘mother’, ModE nun). Ælfric, too, introduced Latin loanwords into his texts. He could draw on words that had been used before him (such as creda, altar, regol; ModE creed, altar, rule), or introduce new ones (such as cranic, pistol, or sanct; ModE chronicle, epistle, saint). In his Grammar he carefully translates every grammatical concept (e.g. nominativus ys nemnendlyc, genitivus is gestrynedlic, dativus ys forgyfendlic, etc.), but despite these translations he uses the Latin terms for the cases throughout. Similarly, he speaks of casus or of dyptongon, and for Modern English declination he uses the OE hybrid word declinung, combining the Lt. declinwith the OE suffix -ung. It is doubtful that many of these learned Latin loanwords introduced by Alfred (or his circle) and Ælfric made it into the vocabulary of the ordinary Anglo-Saxon. Many did not survive: the Alfredian legie has been replaced by legion, tictator by dictator, and the Ælfrician declinung by declination, pistol by epistle, and cranic by chronicle. Nonetheless, during approximately the last 200 years of the Anglo-Saxon period they had a limited currency among the learned.
4 Summary Old English borrowed a total of about 600 words from Latin (Nielsen 1998: 141) of which by necessity only very few could be discussed here. A fuller list is given by Mary Serjeantson (1935) in A History of Foreign Words in English, in the chapter “Latin Words before the Conquest” (11–50) and in “Appendix A” entitled “Pre-Conquest Loan-Words from Latin” (271–288). Though the exact dating of many of these words is disputed, especially Serjeantson’s list of words “probably borrowed in Britain, 450–650”, her Appendix A
371
372
III Old English provides an excellent guide to the various areas in which Latin, whether through direct contact with the Romans or through the book-learning of the clergy, influenced Old English. Overall, the Anglo-Saxons were reluctant to accept Latin loanwords. They preferred to rely on their own linguistic resources, either through infusing a native word with new meaning or through creating a calque, to express the concepts inherent in the Latin words. Old English had an approximate total vocabulary of about 30,000 words (Nielsen 1998: 141); 600 loanwords from Latin, and that over a period of about 800 years, seems a very small proportion indeed. Though Latin might seem to be the language of the superior culture, first as the language of the Roman Empire and then as the language of Christianity, the Anglo-Saxons did not seem overly impressed: they adopted loanwords when they were absolutely necessary, but that necessity was not felt too frequently since Old English possessed the ability to deal with the highly sophisticated concepts of the Latin words through its own Germanic vocabulary.
5 References Bach, Adolf. 1965. Geschichte der deutschen Sprache. 8th edn. Heidelberg: Quelle & Meyer. Bately, Janet (ed.). 1980. The Old English Orosius. (Early English Text Society, S. S. 6.) Oxford: Oxford University Press. Braune, Wilhelm. 1918. Althochdeutsch und Angelsa¨chsisch. Beitra¨ge zur Geschichte der deutschen Sprache und Literatur 43: 398–409. Campbell, A. 1959. Old English Grammar. Oxford: Clarendon Press. Capelle, Torsten. 1998. Die Sachsen des Fru¨hen Mittelalters. Darmstadt: Wissenschaftliche Buchgesellschaft. Capelle, Torsten. 1990. Archa¨ologie der Angelsachsen: Eigensta¨ndigkeit und kontinentale Bindung vom 5. bis 9. Jahrhundert. Darmstadt: Wissenschaftliche Buchgesellschaft. Frank, Roberta. 2001. Old English ancor ‘anchor’: Transformation of a Latin loanword. In: K. E. Olsen, A. Harbus, and T. Hofstra (eds.), Germanic Texts and Latin Models: Medieval Reconstructions. Leuven: Peeters. Fulk, R. D., Robert E. Bjork, and John D. Niles (eds.). 2008 . Klaeber’s Beowulf and the Fight at Finnsburg. Toronto: Toronto University Press. Gelling, Margaret. 1977. Latin loan-words in Old English place-names. Anglo-Saxon England 6: 1–13. Gneuss, Helmut. 1993. Anglicae linguae interpretatio: Language contact, lexical borrowing and glossing in Anglo-Saxon England. In: Proceedings of the British Academy 82, 1992 Lectures and Memoirs, 107–148. Oxford: Oxford University Press. Greenfield, Stanley B. and Daniel G. Calder. 1986. A New Critical History of Old English Literature. New York/London: New York University Press. Jackson, Kenneth. 1953. Language and History in Early Britain. Cambridge, MA: Harvard University Press. Lewis, Charlton T. and Charles Short. 1975. A Latin Dictionary. Oxford: Clarendon Press. Lockwood, W. B. 1976. An Informal History of the German Language. London: Deutsch. Masser, Achim. 1966. Die Bezeichnungen fu¨r das christliche Gotteshaus in der deutschen Sprache des Mittelalters. Berlin: Schmidt. Mitchell, Bruce and Fred C. Robinson. 1992. A Guide to Old English. 5th edn. Oxford/Cambridge, MA: Blackwell. Mu¨ller, Gertraud and Theodor Frings. 1966–1968. Germania Romana II. 2nd edn. Halle: Niemeyer. Myres, J. N. L. 1986. The English Settlements. The Oxford History of England. Vol. 1B. Oxford: Oxford University Press.
24 Old English: Standardization
373
Nielsen, Hans Frede. 1998. The Continental Backgrounds of English and its Insular Development until 1154. Odense: Odense University Press. Pogatscher, Alois. 1888. Zur Lautlehre der griechischen, lateinischen und romanischen Lehnworte im Altenglischen. Strassburg: Tru¨bner. Scha¨ferdiek, Knut. 1984. Kirihha – *cyrica – kyriakon: Zum geschichtlichen Hintergrund einer Etymologie. Beitra¨ge zur Geschichte der deutschen Sprache und Literatur 106: 46–50. Serjeantson, Mary. 1935. A History of Foreign Words in English. London: Routledge and Kegan Paul. Sherley-Price, Leo (trans.). 1993. A History of the English Church and People [by] Bede. Revised edn. New York: Barnes & Noble. Sweet, Henry (ed.). 1871–1872. King Alfred’s West Saxon Version of Gregory’s Pastoral Care. With an English translation, the Latin text, notes, and an introduction. (Early English Text Society 45.) London: N. Tru¨bner & Co. Wollmann, Alfred. 1993. Early Latin loan-words in Old English. Anglo-Saxon England 22: 1–26. Zupitza, Julius (ed.). 2000 [1880]. Aelfrics Grammatik und Glossar. 3rd edn, with a new introduction by Helmut Gneuss. Berlin: Weidmann.
Gernot R. Wieland, Vancouver (Canada)
24 Old English: Standardization 1 2 3 4 5 6 7
Introduction Establishing a norm: earlier scholarly approaches to standardization in Old English Theoretical and descriptive frameworks in current research Major forms of standardization in Old English Further instances of standardization Summary References
Abstract Despite the ongoing debate about the appropriateness of the concept of standardization in an Old English context, scholars concur that the earliest stage of the English language exhibits clear traces of language regulation. Two major processes that differ in their linguistic character and geographical extension have been identified: (1) the so-called “Winchester vocabulary”, a lexical norm taught and practiced at Winchester cathedral school in the late 10th and in the 11th century, and (2) “Standard Old English”, an orthographic norm based on the West-Saxon dialect, whose regulating effect on spelling and inflexional morphology manifests itself in late Old English manuscripts originating in all parts of England from the late 10th to the early 12th century. The sociolinguistic turn brought about by the Norman Conquest deprived the normative tendencies manifest in Old English of their linguistic foundation and their institutional support. The dearth of vernacular sources in early Middle English and the unregulated character of their language highlight the unique position Old English holds among the vernaculars of early medieval Europe as regards its great appreciation as well as its conscious handling and use by the intellectual elites of Anglo-Saxon England. Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 373–385
374
III Old English
1 Introduction If William the Conqueror had not invaded England in the year 1066, standard English would have looked completely different today. Not only would the enormous French component in the English vocabulary have been considerably smaller, the standard language would in all likelihood have had its origin in a different dialect as well.
Introducing the topic of “Standardisation” in the history of English, Terttu Nevalainen and Ingrid Tieken-Boon van Ostade (2006: 271) thus point to the fact that the story of Standard English is markedly discontinuous in chronological as well as in geographical terms. The standardizing process in pre-Conquest times which the above quotation alludes to is subsequently concretized as having “affected the West Saxon dialect, with Winchester as its main cultural centre”. While this statement can be said to be representative of traditional scholarly opinion, Richard Hogg in a handbook article on “Old English Dialectology” published in the same year takes a much more radical stance by concluding “that it is doubtful that there ever existed a variety of West Saxon which could properly be described as forming a ‘Standard Old English’ ” (Hogg 2006: 402). As will be shown in more detail in Section 2, the history of research into standardization – just like its object of study – is by no means straightforward in its development. Current scholarship is still in search of an adequate descriptive framework for the normative processes operative in Old English that is suitable for capturing the differences between the notions of “standardization” and “a standard” in the earliest period of English as compared to its later stages (cf. Section 3). The major forms of standardization in Old English that have been established – the so-called “Winchester vocabulary” and “Standard Old English” – will be dealt with in Section 4. Further instances of standardization that have been identified are briefly addressed in Section 5. The divergences in scholarly interpretation become more understandable if we look at the nature of the available material. Due to its unusually rich and varied transmission of vernacular texts, Old English holds a unique position among the Germanic dialects. Still, the uneven distribution of the manuscript evidence over time and place and the restricted nature of the data make an adequate assessment of regularizing processes in Anglo-Saxon times difficult. Impressive as the vernacular legacy of the Anglo-Saxons may be, it is representative of a small and educationally privileged section of society only. Scholars in search of normative tendencies face in fact quite similar problems to those who look for variation, and it is from historical dialectology that research into standardization in Old English has received both major new impulses and serious challenges during the past few decades.
2 Establishing a norm: earlier scholarly approaches to standardization in Old English The relatively great linguistic homogeneity of the majority of Old English texts preserved in manuscripts dating from c.1000 and later was already recognized by Early Modern English scholars who developed an interest in Anglo-Saxon studies. George Hickes, the author of the first grammar of Old English (1689), based his work (subsequently integrated into his monumental Linguarum Vet[erum] Septentrionalium Thesaurus Grammatico-Criticus et Archæologicus, 1703–5), on the “pure, sweet and
24 Old English: Standardization regular” type of language transmitted by writers in the southwestern parts of AngloSaxon England (“Hactenus de Anglo-Saxonico sermone, quem in auctoribus, qui in australibus & occidentalibus nostræ Britanniæ partibus floruerunt, habemus purum, suavem & regularem, tractavimus”; Hickes 1970 [1703–5] 1: 87). What is now known as “Late West Saxon” lost its status as a reference dialect for the study of Old English when Henry Sweet published his edition of King Alfred’s translation of Pope Gregory’s Cura Pastoralis in 1871. With his claim that West Saxon was “fixed and regulated by the literary labors of Alfred and his successors”, Sweet (1871: xxxii–xxxiii) triggered off “something approaching a revolution in English philology” (Gneuss 1972: 65). Not only was Alfred’s language assigned a normative character and the king seen as the initiator and promoter of an Old English literary standard. Sweet’s didactic concerns had far-reaching consequences for the subject as a whole: the perceived Early West Saxon standard was turned into “the standard for Old English” (Wrenn 1933: 68) and in a normalized form became imported into grammars, dictionaries, and textbooks. A paper by C. L. Wrenn on “ ‘Standard’ Old English”, presented at a meeting of the Philological Society in 1933, served as a major corrective to this development. Wrenn demonstrated that the spellings in the few surviving Alfredian manuscripts from the late 9th and early 10th centuries were too varied and too inconsistent to provide a suitable model for a supraregional standard. Instead, Wrenn pointed to Ælfric’s much more regulated and regular language as a prototypical example of “classical Old English”, “a common and universally used West-Saxon Schriftsprache in the late tenth and early eleventh centuries” (Wrenn 1933: 66, 85). The issue of an Old English standard was taken up again by Helmut Gneuss in a seminal article published in 1972 under the title “The origin of Standard Old English and Æthelwold’s school at Winchester”. Gneuss combined philological expertise with insights from other historical disciplines to provide a more differentiated picture of the late Old English “standard literary language” and to locate the standardizing forces in a specific intellectual and institutional environment. The two normative systems he identified – the so-called “Late West Saxon Schriftsprache” and the “Winchester vocabulary” – were both assigned a common origin: the school established by Bishop Æthelwold (963–984) at his cathedral, the Old Minster in Winchester. Though Gneuss (1972: 81) stressed the hypothetical character of this assumption and the need for further research, uncritical reception of this much-quoted publication has led to a frequent confounding of the two norms, which differ in linguistic character, geographical reach and purpose (cf. Gneuss 1972/1996: Addenda).
3 Theoretical and descriptive frameworks in current research Standardization is the stepchild of Old English grammar writing, which, for obvious reasons, has primarily sought to document the full amount of dialectal diversity in vernacular texts. Thus standardizing processes have usually not been traced in a systematic way and statements on the nature of particular “standards” are few and far between. Alistair Campbell (1959: 7–8), for instance, in his Old English Grammar explicitly acknowledges the existence of an Old English standard, stating that “[a]fter 900 the use of West-Saxon as a standard language reduced the writing of Mercian”. Subsequently, however, he takes a much more cautious stance: “Even when West-Saxon had become a
375
376
III Old English well-established literary dialect, and was used as something of a standard written language, many manuscripts display a considerable non-West-Saxon element in their orthography and inflexions” (Campbell 1959: 9 [my italics]). In the light of his later position, Richard Hogg in his Grammar of Old English describes the emergence of an Old English standard in remarkably unequivocal terms: there begins to emerge in the latter part of the tenth century a written standard language or Schriftsprache with a stable orthographic system. The Schriftsprache is most obviously associated with the works of Ælfric, […], but is more generally found, and it may be taken as the basis of a standard or classical OE, extending for about the last hundred years before the Norman Conquest (Hogg 1992: 3).
Hogg’s later doubts “about the usefulness of the concept of standardization in the Old English context” (Hogg 2006: 414) resulted from a critical examination of the terminological foundations of Old English dialectology. Further theoretical and methodological impulses have come from scholars taking a wider diachronic perspective, who are inevitably confronted with the problem that the various “standards” identified in the history of English are only comparable to a limited extent. Major attempts to increase the explanatory potential of the terminology involve, for example, the establishment of a distinction between “standard” or “fixed” and “standardized” or “focused” languages, though in practice these categories can be so differently conceptualized that they turn out to be incompatible. Norman Blake (1996: 7–8), for example, defines the difference between “standard” and “standardized” as “largely political and educational”: “A standardized language” – i.e., “a language which has achieved a reasonable measure of regularity in its written form” – “remains either regional or personal; a standard language has been adopted widely throughout the country”. By contrast, Jeremy Smith’s distinction between “standard/fixed” and “standardized/focused” codes is based on linguistic criteria, with medieval written standards being understood as “a sort of mean towards which scribes tend”. Smith (1996: 67) suggests that in pre-modern times “we are dealing with a process of normative focusing rather than with a fixed set of forms”. His concept of a “standardized” or “focused” norm allows for much more internal variation than the traditional notion of “a standard” is usually thought to permit. Its prestige may make the standardized variety the focus of attraction for other varieties and may thus function as a norm-enforcing mechanism. There is also no inbuilt tendency that necessarily turns a “standardized” into a “standard” variety. The state of “fixity” and normative rigidity, which forms a defining characteristic of Smith’s concept of a “standard language”, is absent from Blake’s scheme. Here the chief locus of linguistic regularization is the “standardized language”, while the “standard language”, whose “characterizing feature is the political and educational will to impose a standard on the country as a whole” (Blake 1996: 8) may exhibit varying degrees of normative regularity. He therefore regards the development of a “standard language” into a “standardized language” as a natural process, which, however, does not work in reverse (Blake 1996: 7). Such fundamental differences in the understanding of linguistic core concepts inevitably lead to different, at times even contradictory, results in the classification and evaluation of normative developments in Anglo-Saxon England. Further attempts to take a more theoretically informed approach to standardization in Old English include the application of Haugen’s classical four-stage model to
24 Old English: Standardization “Ælfrician English” (cf. Hogg 2006) and the explanation of the regularizing activities of the “Winchester School” in terms of social network theory (cf. Lenker 2000). While the first and the last step in Haugen’s (1966: 933) scheme – (1) selection of norm, (2) codification of form, (3) elaboration of function, and (4) acceptance by the community – acknowledge the role of society in the construction of a standard, social-network approaches to language change take the nature and strength of community ties as determining factors for the creation, enforcement, and maintenance of linguistic norms. Following Labov (1994: 78), the standardization of Old English has also been described as a conscious “change from above” and, in James and Lesley Milroy’s terms (Milroy and Milroy 1999: 6, 22), as an instance of “suppression of optional variability” resulting from “a need for uniformity that is felt by influential portions of society at a given time” (cf. Nevalainen and Tieken-Boon van Ostade 2006: 272–273). In practice, though, the systematic testing of the Milroys’ extended model, which distinguishes seven stages of implementation, has been reserved for the standardizing processes starting in Late Middle English (cf. Nevalainen and Tieken-Boon van Ostade 2006: 272–286). It seems clear that we cannot judge the degree of standardization in Old English by applying modern standards. The search for an appropriate mode of classification does, however, raise important questions: If standardization means “intolerance of optional variability” (Milroy and Milroy 1999: 22), how much tolerance can, or perhaps even must, be built into the standard classificatory systems to accommodate the conditions of a distant period? It is hardly possible to determine exactly how much variation would still have been regarded as “normal” by a norm-conscious Anglo-Saxon writer or any of the scribes and copyists trained in a specific normative system. Yet on the basis of new textual editions which offer material that was not available to former editors (and usually also not to the authors of the major grammars of Old English), and with the help of searchable digitalized corpora, a more detailed and more informed picture of linguistic normativity in Anglo-Saxon England can emerge.
4 Major forms of standardization in Old English 4.1 The historical and cultural background There is a general consensus that a successful standardization process requires a sufficient amount of institutional support. Most scholars have followed Gneuss (1972) in locating the Anglo-Saxon norm-setting elites primarily in ecclesiastical and monastic circles backed by monarchic governments. The exact role which the chancery – “the royal writing office of the tenth and eleventh centuries, staffed by priests who served in the royal household” (Keynes 1999: 94) – may have taken in this context is not clear; as the production of charters (usually in Latin) and writs (in the vernacular) lay in ecclesiastical hands, the chancery scribes in all likelihood received their training also in religious houses. Though references to King Alfred and his program of translation are still common in textbooks on the history of English, most scholars agree that both on linguistic grounds and on the basis of other historical evidence the King cannot be assigned a key role in the standardization of Old English. As “king of the Anglo-Saxons” Alfred no doubt “provided the springboard for his successors of the tenth century to become kings of England” (Yorke 1999: 28). There is, however, still no proof that his literary activities (on which see the sceptical remarks by Godden 2007 and the rejoinder by
377
378
III Old English Bately 2009) “raised West Saxon above the status of a dialect” (Gneuss 1972: 68). Thus, Blake’s claim that the dissemination of the Old English translation of the Pastoral Care throughout Anglo-Saxon England “marks the start of a standard English, even if the form it took was not at first highly standardised” (Blake 1996: 86) remains highly controversial. The decisive impetus for the main regulative processes affecting Old English came from the 10th-century revival of monasticism, culture, and learning known as the “Benedictine Reform”. As suggested by Gneuss (1972), Bishop Æthelwold’s cathedral school in the royal capital Winchester is still regarded as a focal place for the cultivation of linguistic norms; recent research has, however, provided a more differentiated picture of the political and cultural environments in which such normative ideas could unfold. In a number of publications, Mechthild Gretsch (cf. especially 1999, 2001) has demonstrated that the formation of the “Kingdom of the English” during the 10th century and the ideological principles guiding the West Saxon rule over all England were essential for the monastic zeal for uniformity to take effect. Gretsch ascribes the origins of the “Winchester vocabulary” to the formative years of Benedictine reformed monasticism when we find the chief protagonists of the religious reform movement, Æthelwold and Dunstan (the later archbishop of Canterbury, 959–988) at the court of King Æthelstan (reigned 924–939) and in joint study at Glastonbury Abbey (c.939–954). The orthographic standardization of Old English is assigned to the concluding years of King Edgar’s reign (959–975), a period characterized by a close cooperation between the monarch and the leading monastics, and by strong normative tendencies on both sides. The early 970s saw such important regulative acts as the introduction of a new currency by the king (“King Edgar’s reform of coinage”) and the promulgation of the Regularis concordia – a monastic customary that was to secure a uniform observance in Benedictine houses throughout the country. The reformers’ interest in standardizing the use of the written medium, which will be discussed in the following sections, also manifests itself in a systematic use of two different types of scripts: Anglo-Caroline minuscule for Latin texts and Anglo-Saxon minuscule for texts in the vernacular (see the brief overview in Gretsch 2003: 36–39).
4.2 The “Winchester vocabulary” The most prominent example of lexical standardization in Old English is the so-called “Winchester vocabulary”, which Gneuss (1972: 78) described as “a specific and planned vocabulary, prevalent in one school and restricted to a certain area” (see also Sauer and Waxenberger, Chapter 22). As the late 10th- and early 11th-century texts in which this lexical norm can be traced are all in some way connected to Winchester, he dubbed them “the Winchester group” (1972: 76). Its core component is the work of Ælfric of Eynsham (c.950–c.1010), the most outstanding product of the famous cathedral school, who in various contexts defined himself as “alumnus Æthelwoldi” ‘a pupil of Æthelwold’, or “Wintoniensis alumnus” ‘a pupil of Winchester’ (cf. Wilcox 1994: 7). As Lenker (2000: 238) argues, the fixation of lexical choices for certain key concepts by the “Winchester circle” makes the “Winchester vocabulary” “a model case of cultural and linguistic focusing in a tightknit network”. In an extensive study, Walter Hofstetter (1987; conveniently summarized in Hofstetter 1988) investigated the degree to which this regulated vocabulary permeates Old
24 Old English: Standardization English literature. His testing material consisted of thirteen word groups from different semantic fields, with each word group comprising three types of “synonyms”, i.e. lexical items “which can, in certain contexts, fulfil the same semantic function” (Hofstetter 1988: 143): “A words”, reflecting Winchester usage, “C words”, which were avoided there, and “B words”, which show an indistinctive distribution and therefore proved unsuitable for classifying the relevant texts. Thus, for ‘virtue’ in the religious or moral sense (Lt. virtus) the Winchester word (A) is miht, the non-Winchester words (C) are cræft, mægenðrymm, strengð, and strengu, and the “B word” occurring in texts with either affiliation is mægen. In sum, the study yielded two large groups of texts (see Hofstetter 1988: 151–156) which by their marked preference for or avoidance of “Winchester words” were apt to prove the existence of this lexical norm and to define the limits of its sphere of influence in a more precise way. It seems not surprising that Ælfric’s work – the norm for the norm – scored highest among the texts characterized by Winchester usage (98.3%). What begs for an explanation, however, is the relatively modest rating of the main Old English work produced by his revered teacher – Æthelwold’s translation of the Benedictine Rule (62.1%). Hofstetter’s (1988: 142, 157) argument that this text as well as the Old English interlinear gloss to the Royal Psalter – another work associated with Æthelwold – show the Winchester usage in statu nascendi has been expounded by Gretsch (1999, 2001: 45–46), who sees the leading reformer experimenting with the relevant lexis already during his studies at Glastonbury in the 940s or early 950s. Confirmation of a regularized lexical usage linked to Winchester has also come from studies in Old English word geography such as Elmar Seebold’s (1974) examination of the Old English equivalents of Lt. sapiens and prudens. One of the four groups of Old English texts of southern provenance which Seebold established, the so-called “Benediktiner-Gruppe”, corresponds to Gneuss’s “Winchester group”. In contrast to Gneuss (1972: 76), who ascribed the distinctive choice of vocabulary discernible in these texts to “stylistic considerations”, Seebold (1974: 330–331) interpreted the lexical divergences between his “Benedictine group” and the other three groups as manifestations of particular South English subdialects. The question if and to what extent the “Winchester vocabulary” is rooted in specific local or regional varieties is still a disputed matter. Peter Kitson, whose dialect studies are based mainly on an analysis of vernacular charter boundaries, has repeatedly pointed out that “ ‘Winchester usage’ in the words covered by it cuts across ordinary dialect distributions” (Kitson 1995: 103 n. 20). The current state of scholarship suggests that there is probably no single answer to the question about the origin of the “Winchester vocabulary”. In fact the proponents of “Winchester usage” seem to have drawn on various types of lexical material – including genuine dialect terms such as oga ‘terror’ – and adapted it to their needs. Most of the “Winchester words” occur in religious writings; they can thus be assigned to a theological register or show specific usages in accordance with Christian doctrine (cf. e.g. the nominal concepts ‘martyr’ and ‘martyrdom’, ‘virtue’, and ‘pride’, and verbal concepts such as ‘to correct’ and ‘to repent’). Even though no uniform picture about the character of the “Winchester vocabulary” has emerged so far, research has revealed a number of principles that appear to have guided the selection, formation, and appropriation of particular items. As Gretsch (2001: 57) has pointed out, it is essential to keep in mind that Winchester usage was “introduced and established on the foil of Latin”. The “Winchester group”
379
380
III Old English consists primarily of prose translations of Latin texts or Old English interlinear glosses. The teaching of standard Latin-Old English equivalents and the practice of glossing exercised at Æthelwold’s school in Winchester may thus have exerted a considerable influence on the fixation and enforcement of this educated usage. In a fashion characteristic for Old English translation procedures, the “Winchester vocabulary” extends into the field of word-formation. Three examples may suffice to demonstrate the principles of this regulated usage, which in a number of cases turns out to be much more specific than the polysemous Latin models. Gelaðung ‘church = the whole Christian community’ (derived from OE laðian ‘to invite, summon’) and cyþere ‘martyr’ (derived from OE cyþan ‘to proclaim, testify, confess’) can be classified as “etymologizing translations” which – in accordance with current glossematic techniques – bring out the original meaning of the lemmata they render (Lat. ecclesia and martyr; cf. Gretsch 2001: 54–55). In addition, these neologisms were probably coined with “pastoral considerations” in mind: in employing such transparent formations “the reformers probably wanted to express concepts central to the Christian religion more vividly by the use of native terms than could be done with the corresponding loanwords cirice and martir” (Hofstetter 1988: 160). A conscious creative act of a different type can be assumed in the case of wuldorbeag (wuldor ‘glory’ + beag ‘ring’), which renders Lt. corona in the metaphorical sense ‘crown of glory’. The usage of this term not only secured a greater degree of semantic precision than the alternative terms beag (a “C word”) and cynehelm (a “B word”), which were also employed in a secular sense (Hofstetter 1988: 160). According to Gretsch (2001: 58, 66–68) this “flamboyant coinage” also echoes Æthelwold’s predilection for the Latin “hermeneutic style” and the spirit of the time. Gretsch’s plea for an augmentation of the acknowledged set of “Winchester words” by other words “revealing the philological preoccupations of Æthelwold’s school” (Gretsch 2001: 48) has led her into the field of technical terminologies. Her study embraces the language of liturgy, whose use is attested in “Glastonbury-Winchester circles”, and Ælfric’s grammatical terminology, which can be shown to have gained some currency outside Winchester circles (Gretsch 2001: 48–53). Though consistency in use was no doubt an important aim for the practitioners of Winchester usage, their “concern for style” and “elegant variation” seems to have admitted of a greater freedom of choice than traditional notions about the rigidity of this norm usually concede. Such motivated “deviations from the Winchester standard” (Hofstetter 1988: 161 n. 113) have especially been demonstrated for Ælfric (see also Gretsch 2001: 46, 51, 63, with further bibliographical references).
4.3 The Late West Saxon Schriftsprache and the concept of “Standard Old English” Though “Standard Old English” and the “Winchester vocabulary” share some common traits as regards their institutional foundations, their assumed promoters and their underlying motivation, they clearly differ in linguistic character and scope. The locus classicus for successful standardization in past ages was – and still is – spelling. “Standard Old English” is commonly defined as an orthographic norm based on the Late West Saxon dialect, but attested in late 10th, and especially 11th century English manuscripts throughout the country. Its designation as a “Schriftsprache” implies that no claim can be made that this formal written code also gained a wider currency as a
24 Old English: Standardization spoken prestige norm. As the conservative character of the standardized spellings masked changes in pronunciation, the distance between symbols and sounds must have continuously increased even in the dialectal home of the standard. The distinctive features of “Standard Old English” have chiefly been located in the graphemic representation of stressed vowels and in morphosyntax: the preservation in writing of full vowels in inflexional endings largely conceals the phonetic reduction of unaccented vowels to schwa /ə/. As deviant spellings reveal, this levelling process must have been fairly progressed in late Old English speech. It is now commonly agreed that the highly regulated orthography that shows up in late Old English texts is no direct continuation of what has variously been termed “Early West Saxon” or “Alfredian Old English”. Hogg (2006: 402) has tentatively assigned this variety the status of an “earlier, but ultimately less successful, focused language”. The doubtful nature of the term “Alfredian Old English” from a norm-oriented perspective has been demonstrated by Gretsch (2000). A substantial part of the small “Alfredian” corpus of four manuscripts represents a later stage of transmission, approximately from the 920s, and shows scribal and linguistic links to the Old English gloss to the Junius Psalter, itself a “West-Saxonized” version of the Mercian Vespasian Psalter gloss. Gretsch (2001: 77) postulates that “by the time of King Edward the Elder (899–924) a type of literary language had developed which, although basically West Saxon, tolerated Anglian forms and words, and would thus have been England’s first supradialectal language, reflecting the political order of its time, the Kingdom of the Anglo-Saxons”. In a similar vein, the relationship between “Late West Saxon”, “Standard Old English”, and “Ælfrician Old English” has recently come under closer scrutiny, not least because an indiscriminate use of these concepts carries the danger of circular reasoning. There is a general consensus that, like all Old English dialect labels, “Late West Saxon” is an abstraction which has to allow for a considerable degree of internal variation. As has already been pointed out in Section 3, this is to some extent also true for its written form, which is frequently equated with “Standard Old English” without further qualification. “Standard Old English” in turn becomes personalized in the term “Ælfrician Old English”, because the prolific Old English prose writer is commonly regarded as a model practitioner of the Late West Saxon Schriftsprache. Doubts about the status of “Ælfrician Old English” as a standard language may thus lead to quite radical conclusions: after applying Haugen’s (1966) criteria for a fully developed standard to Ælfric’s language (albeit in a rather cursory way), Hogg (2006: 402) calls into question the whole idea of a “Standard Old English” that grew out of a particular variety of West Saxon (cf. quotation in Section 1). In his opinion, Ælfrician Old English only fulfils Haugen’s first criterion – selection and use in an important center (the Old Minster at Winchester) – but fails to meet the other essential requirements of a standard, i.e., codification by some external authority, elaboration by extending into new, in this case non-religious, areas, and nationwide acceptance (Hogg 2006: 401). Though one may prefer to classify Ælfric’s writings as attestations of “a focused language rather than a standard language” (Hogg 2006: 401), the outstanding regularity of his usage and his great concern for linguistic correctness are beyond question. There is proof that Ælfric supervised the production of copies of his two series of Catholic Homilies (cf. Gretsch 2003: 41–42, with bibliographical references). His own correcting hand has been identified in the earliest surviving manuscript of the First Series
381
382
III Old English (London, British Library, Royal 7. C. xii). Summarizing the results of her examination of the inflexional forms, Connie Eble (1970: 85) describes “the West Saxon language in Royal” as “regular, conservative, and to some extent artificial – all characteristics generally associated with standard languages”. To what extent Ælfric himself and the manuscripts of his works came to function as linguistic norm-setters has still to be researched in detail. In this context we must not underrate the norm-enforcing influence exerted by his Latin-Old English Grammar (ed. Zupitza 2000 [1880]). Judging by the number of surviving manuscripts, there was at least one copy available in every library in 11th century England, and the text remained in active use for some time after the Norman Conquest (Gneuss 1996: 11). Inflexional morphology – a core field of orthographic standardization in late Old English – is the centre of grammatical attention in this standard textbook. As with the “Winchester vocabulary”, the prestigious model of Latin as a highly standardized written language has probably served as an important source of inspiration for Anglo-Saxon scholars to regularize their vernacular (cf. Gretsch 2001: 76–77; 2003: 60). Recent systematic research into Ælfric’s own forms as compared to scribal forms in Ælfrician manuscripts in and outside his sphere of influence has yielded valuable information about individual spelling practices and competing orthographic norms. Examining manuscript variants recorded in the electronic “Inventory of Script and Spellings in Eleventh-Century English” (http://www.arts.manchester.ac.uk/mancass/C11database), Scragg (2006: 185) concludes that “Ælfric’s scribes, although for the most part very consistent in their copying, are not necessarily transmitting his spellings”. Two further pilot studies into the linguistic transmission of the Lives of Saints and a selected example from the Catholic Homilies conducted by Gretsch (2003, 2006) provide some indications that “inflexional morphology was considered the more important branch of Standard Old English” (Gretsch 2003: 60) and show Ælfric in a number of cases experimenting in search of a stable norm. As regards the relationship of “Ælfrician Old English” and “Standard Old English”, Gretsch (2006: 172) tentatively assumes that “what Ælfric wrote was not ‘Standard Old English’ per se, but ‘Ælfric’s Standard Old English’, and that this existed side by side with other standards, though perhaps none as systematic as his was”. “Standard Old English per se” turns out to be a surprisingly under-researched subject. We still lack a comprehensive survey of the linguistic forms attested in late 10th and 11th century Old English manuscripts that could tell us more about patterns of normative adherence and margins of tolerance in texts that were newly produced or copied from older models in particular scriptoria or by individual scribes. Gretsch’s survey on “Standard Old English and its Acceptance” in a number of non-Ælfrician manuscripts (Gretsch 2001: 69–75) reveals a varying degree of conformity to the supposed standard. That Late West Saxon served as a centripetal prestige norm that attracted users of other varieties can, for example, be demonstrated by the attempts of Farman, the late 10th century Mercian glossator of part of the Rushworth Gospels (Ru1), to adapt his own usage to West Saxon forms, which resulted in a number of hyperadaptations (cf. Smith 1996: 26–29). Texts in the Late West Saxon standard orthography continued to be copied after the Norman Conquest well into the 12th century – often very faithfully, as for instance the annals for 1070–1121 in the Peterborough Chronicle show (cf. Gretsch 2001: 71–72). However, without the necessary institutional support and cultural grounding this “artificial” norm was doomed to die.
24 Old English: Standardization
5 Further instances of standardization Besides the major processes outlined above, scholars have identified a number of nonWest Saxon writing traditions connected with certain dialect areas, institutions, or manuscript groups. Alistair Campbell (1959: 11) has emphasized the difficulties in tracing such local norms, because even the evidence we have from known centers can be quite contradictory. He notes, for example, a “steady tendency towards the development of a local Schriftsprache” in the 9th century Kentish charters, while later documents from this area are often characterized by “the wide use of a south-eastern koine¯, which had gained prestige by use at Canterbury” (1959: 11). Specifically referring to lexical conventions, Elmar Seebold (1989: 59–60) finds the “church language” (“Kirchensprache”) of Canterbury characterized by a conscious attempt to avoid the use of regionalisms in order to secure general understandability – quite in contrast to Winchester, which tended to promote local usage. The most enduring of the locally restricted norms that have been postulated by scholars is the so-called “Mercian literary language”. It has been traced in the 9th century Vespasian Psalter gloss, in the post-Conquest Life of St Chad, and – as a later Middle English reflex – in the 13th century AB-language. Hogg (2006: 404) describes it as a “focused language” that represents “an attempt to rationalize spelling traditions to a much greater extent than elsewhere” (for the controversial discussion of its normative status and temporal extension see Schaefer, Chapter 33). A special case of regulated usage is the language of Old English poetry, which came under the standardizing influence of Late West Saxon in the process of transliteration around the turn of the millennium. This specialized language is “relatively homogeneous” (Godden 1992: 491) in its mixed character. Whether this justifies the assumption of a “general Old English poetic dialect” is a moot point (see Fulk, Chapter 25).
6 Summary Despite its long and impressive scholarly history the subject of standardization in Old English still calls for further work on the theoretical and on the practical side to define the exact nature of the various “standards” that have been identified, their dissemination and their acceptance. Investigations into the details of regulated usage on the basis of an extended range of data have confirmed the existence of normative systems, while at the same time demonstrating the need for a less rigid conceptualization of their homogeneity and the consistency of their application. As Richard Hogg (2006: 414) pointed out, “it is quite possible that what we are witnessing in the West Saxon area during the tenth century is something more complex but also more interesting than has been previously thought”.
7 References Bately, Janet. 2009. Did King Alfred actually translate anything? The integrity of the Alfredian canon revisited. Medium Ævum 78: 189–215. Blake, N. F. 1996. A History of the English Language. Houndmills: Macmillan. Campbell, A. 1959. Old English Grammar. Oxford: Clarendon.
383
384
III Old English Eble, Connie. 1970. Noun Inflection in Royal 7.C.XII, Ælfric’s First Series of Catholic Homilies. Unpublished Ph.D. dissertation, University of North Carolina at Chapel Hill. Gneuss, Helmut. 1972. The origin of Standard Old English and Æthelwold’s school at Winchester. Anglo-Saxon England 1: 63–83. (Reprinted with Addenda in: Helmut Gneuss, Language and History in Early England, no. I. Aldershot: Ashgate, 1996.) Gneuss, Helmut. 1996. English Language Scholarship: A Survey and Bibliography from the Beginnings to the End of the Nineteenth Century. Binghamton, NY: Center for Medieval and Early Renaissance Studies. Godden, Malcolm. 1992. Literary language. In: Richard M. Hogg (ed.), The Cambridge History of the English Language. Vol. I. The Beginnings to 1066, 490–535. Cambridge: Cambridge University Press. Godden, Malcolm. 2007. Did King Alfred write anything? Medium Ævum 76: 1–23. Gretsch, Mechthild. 1999. The Intellectual Foundations of the English Benedictine Reform. Cambridge: Cambridge University Press. Gretsch, Mechthild. 2000. The Junius Psalter gloss: Its historical and intellectual context. AngloSaxon England 29: 85–121. Gretsch, Mechthild. 2001. Winchester vocabulary and Standard Old English: The vernacular in Late Anglo-Saxon England. The T. N. Toller Memorial Lecture 2000. Bulletin of the John Rylands University Library of Manchester 83: 41–87. Gretsch, Mechthild. 2003. In search of Standard Old English. In: Lucia Kornexl and Ursula Lenker (eds.), Bookmarks from the Past. Studies in Early English Language and Literature in Honour of Helmut Gneuss, 33–67. Munich: Fink. Gretsch, Mechthild. 2006. A key to Ælfric’s Standard Old English. In: Swan (ed.), 161–177. Haugen, Einar. 1966. Dialect, language, nation. American Anthropologist 68: 922–935. Hickes, George. 1970 [1703–5]. Linguarum Vet[erum] Septentrionalium Thesaurus GrammaticoCriticus et Archæologicus. (2 volumes in 1.) Hildesheim: Olms. (First published Oxford: Theatrum Sheldonianum.) Hofstetter, Walter. 1987. Winchester und der spa¨taltenglische Sprachgebrauch. Untersuchungen zur geographischen und zeitlichen Verbreitung altenglischer Synonyme. Munich: Fink. Hofstetter, Walter. 1988. Winchester and the standardization of Old English vocabulary. AngloSaxon England 17: 139–161. Hogg, Richard M. 1992. A Grammar of Old English. Vol. 1. Phonology. Oxford: Blackwell. Hogg, Richard. 2006. Old English Dialectology. In: Ans van Kemenade and Bettelou Los (eds.), The Handbook of the History of English, 395–416. Oxford: Blackwell. Keynes, Simon. 1999. Chancery, royal. In: Lapidge et al. (eds.), 94–95. Kitson, Peter. 1995. The nature of Old English dialect distributions, mainly as exhibited in charter boundaries. Part I: Vocabulary. In: Jacek Fisiak (ed.), Medieval Dialectology, 43–135. Berlin/ New York: Mouton de Gruyter. Labov, William. 1994. Principles of Linguistic Change. Vol. 1. Internal Factors. Oxford/Malden, MA: Blackwell. Lapidge, Michael, John Blair, Simon Keynes, and Donald Scragg (eds.). 1999. The Blackwell Encyclopaedia of Anglo-Saxon England. Oxford: Blackwell. Lenker, Ursula. 2000. The monasteries of the Benedictine Reform and the “Winchester School”: Model cases of social networks in Anglo-Saxon England? European Journal of English Studies 4: 225–238. Milroy, James and Lesley Milroy. 1999. Authority in Language. Investigating Standard English. 3rd edn. London/New York: Routledge. Nevalainen, Terttu and Ingrid Tieken-Boon van Ostade. 2006. Standardisation. In: Richard Hogg and David Denison (eds.), A History of the English Language, 271–311. Cambridge: Cambridge University Press. Scragg, Donald. 2006. Ælfric’s scribes. In: Swan (ed.), 179–189. Seebold, Elmar. 1974. Die ae. Entsprechungen von lat. sapiens und prudens: Eine Untersuchung u¨ber die mundartliche Gliederung der ae. Literatur. Anglia 92: 291–333.
25 Old English: Literary language
385
Seebold, Elmar. 1989. Winchester und Canterbury: Zum spa¨taltenglischen Sprachgebrauch. Anglia 107: 52–60. Smith, Jeremy J. 1996. An Historical Study of English: Function, Form and Change. London: Routledge. Swan, Mary (ed.). 2006. Essays for Joyce Hill on her Sixtieth Birthday. Leeds Studies in English 37. Leeds: School of English, University of Leeds. Sweet, Henry (ed.). 1871. King Alfred’s West-Saxon Version of Gregory’s Pastoral Care. (Early English Text Society, OS 45, 50.) London: Oxford University Press. Wilcox, Jonathan (ed.). 1994. Ælfric’s Prefaces. Durham: Department of English Studies. Wrenn, C. L. 1933. “Standard” Old English. Transactions of the Philological Society 32: 65–88. Yorke, B. A. E. 1999. Alfred. In: Lapidge et al. (eds.), 27–28. Zupitza, Julius (ed.). 2000 [1880]. Aelfrics Grammatik und Glossar. 3rd edn., with a new introduction by Helmut Gneuss. Hildesheim: Olms-Weidmann. (First published Berlin: Weidmann.)
Lucia Kornexl, Rostock (Germany)
25 Old English: Literary language 1 2 3 4 5
Introduction The language of poetry The language of literary prose Summary References
Abstract Old English as it is preserved represents a variety of literary standards in competition, often within a single text. A range of registers is evident in both poetry and prose, and many of the features characteristic of an elevated style are to be associated with the Anglian dialects (Mercian and Northumbrian), due to the political and cultural ascendancy of the Anglian kingdoms in the early portion of the period. The language of nearly all of the preserved poetry is a koine, chiefly West Saxon in character but with a strong admixture of orthographic, morphological, and syntactic features from other dialects. In varying degrees, such nonstandard characteristics may also be found in prose. Lexis is keyed to register, as well, there being many items of strictly poetic vocabulary (some of them occasionally used for rhetorical effect in prose) and a smaller number of strictly prosaic words. Figures of rhetoric, some Latinate, are frequent in poetry, and in prose they are especially common in homilies.
1 Introduction A remarkable range of registers is detectable in the language of Old English literature. Even when discontinuous discourse, such as various sorts of lists and glosses on Latin texts, is excluded, the language of prose varies widely in its linguistic features (and at multiple levels of analysis), from the telegraphic and obscure expressions of legal and Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 385–398
386
III Old English penitential literature to the complex, hypotactic structures of some Alfredian prose. The elevated features of poetic literary language are far more distinctive than those of prose. Yet in poetry, too, there is an obvious range of registers, from the relatively prosaic syntax and lexis of the Alfredian Meters of Boethius, along with a great deal of later verse, to the intricate periods of poems composed in the classical manner, i.e. in the most traditional and archaic-seeming style, especially Beowulf. A consideration that complicates the analysis of literary language is that a large portion of the Old English corpus is translated from Latin. This complication is of greater significance in regard to prose, since poetic translations tend to be freer. But even in poetry, certain syntactic patterns are likely to be due to the influence of Latin models. No Latin model is required, however, for the marked differences between the language of verse and of prose, as many of the relevant features of Old English poetic discourse are observable also in languages cognate with Old English.
2 The language of poetry There exist four codices, informally known as the Junius Manuscript, the Vercelli Book, the Exeter Book, and the Beowulf Manuscript, in which most of the approximately 30,000 surviving lines of Old English poetry are preserved. The poems range in length from 2 to 3,181 lines, in date of composition from the 7th century to the early 12th (though few poems can be dated securely), in quality of execution from doggerel to the sublime, and in subject across a range of categories including lyric, liturgy, medical charms, saints’ lives, scripture, philosophy, proverbs, history, and heroic legend. Nearly all the poetry is preserved in a variety of the Late West Saxon dialect that displays a strong admixture of features normally encountered in other dialects, particularly Anglian (Mercian and Northumbrian). Almost certainly, most of the surviving poems were composed in Anglian (Angl.) and subsequently rendered into West Saxon (WSax.). But West Saxon scribes plainly allowed Anglian forms in poetry that they did not allow in prose, and even some poems known to have been composed in Wessex, including those by King Alfred the Great, contain some Anglian forms. Therefore, it is generally assumed that the language of verse is a koine, a poetic dialect native to no one place, like Homeric Greek (Sisam 1953: 119–139). The reason for the admixture of particularly Anglian elements is the greater prestige and association with ancient traditions that attached to Anglian linguistic forms, since Northumbria and Mercia were the chief centers of power and culture in the early years of the Anglo-Saxon period. Accordingly, Mercian influence is also apparent in West Saxon and Kentish prose of the 9th century (A. Campbell 1969: 9–11). (See further Sauer and Waxenberger, Chapter 22.)
2.1 Phonology Examples of Anglian features in the poetic koine are the following: (1) retraction (rather than breaking) of æ before covered (i.e., anteconsonantal) l (as in cald, WSax. ceald ‘cold’); (2) æ as the front mutation of the vowel thus produced (as in wælm, Early WSax. [EWSax.] wielm, Late WSax. [LWSax.] wylm ‘fervor’); (3) West Mercian (and Kentish) raising of æ to e (as in meþel, WSax. mæþel ‘council’);
25 Old English: Literary language (4) Anglian (and Kentish) e as the front mutation of ea before covered r (as in ermðu, EWSax. iermðu, LWSax. yrmðu ‘misery’); (5) Anglian (and Kentish) e¯ as the reflex of Germanic e¯1 (as in de¯d, WSax. dæ¯d ‘deed’); (6) Anglian (and Kentish) e¯ as the front mutation of e¯a (as in ge¯man, EWSax. gı¯eman, LWSax. gy¯man ‘observe’); (7) Anglian (and Kentish) failure of diphthongization by initial palatal consonant (as in gelp, EWSax. gielp, LWSax. gylp ‘boast’); and (8) breaking in seolf ‘self’ (WSax. self, silf, sylf ). Distinctively Kentish are: (9) e as the front mutation of æ broken to ea before covered l (e.g. in elde, EWSax. ielde, LWSax. ylde, Angl. ælde ‘men’); (10) graphic confusion of e¯̆ and y¯̆ due to their merger (as in senne, elsewhere synne ‘sin’); (11) eo as the front mutation of io before covered r (as in beorhto, EWSax. bierhtu, LWSax. byrhto, Angl. birhtu ‘brightness’); (12) ¯ı̆ o for e¯̆ o (as in wiorðan, WSax. weorðan ‘become’). Almost all of the phonological dialect features of the koine are also found in poems of West Saxon as well as of Anglian composition (see Fulk 1992: 283–308, with references). They are, however, more common in poems generally regarded as Anglian in origin. This is particularly true of Anglian smoothing, which monophthongized back diphthongs before (presumably) palatal consonants (as in ferh, WSax. feorh ‘life’). It is also true of back mutation (diphthongization of a front vowel due to the appearance of a back vowel in the following syllable), which in verse occurs in many more environments than it does in West Saxon prose (as in riodon, WSax. ridon PL ‘rode’). Phonology naturally plays a fundamental role in the construction of verse form, to which alliteration is essential. The poetic line comprises a pair of verses, in which the onset of the first fully stressed syllable in the first verse (and, in many types, a second such onset) is identical to that of the first fully stressed onset in the second verse (in which a second alliterating onset is disallowed). Under the rules of alliteration, any vowel alliterates with any other vowel, while a consonant alliterates only with an identical consonant. An exception is that [j] and [ɡ] (both spelt and deriving, in part, from [ɣ]) are matched, as are [tʃ ] and [k] (both spelt and deriving from [k]). It is only in compositions of the second half of the 10th century and later that the different sounds represented by identical orthographic symbols are no longer consistently matched: in The Battle of Maldon (991 or later), for example, [j] alliterates only with [j], and [ɡ] only with [ɡ]. Furthermore, in poetry of all dates, each of the graphemic clusters (= [ ʃ ]), , and alliterates only with itself, and not with alone (see Minkova 2003). Prosody also plays a key role in the construction of verse form. Under the most widely credited analysis of the poetic meter (Sievers 1885, 1893, and other approaches based on Sievers’s views; for a concise account, see Pope 2001: 129–158), at least three levels of stress play a role (full stress, half-stress, non-stress), and syllable quantity is also crucial. In respect to prosody, a salient feature of the language of poetry is the extent to which meter indicates the preservation of archaic phonological patterns.
387
388
III Old English Thus, for example, a verb like se¯on ‘see’, from earlier *seohan, may demand scansion as two syllables as late as the end of the 9th century, though loss of h and subsequent vowel contraction must have occurred much earlier. Similarly, a word like hleahtor ‘laughter’ may require scansion as a monosyllable, as if it had not undergone syllabification of the final resonant, a process that (as regards r, at least) antedated any extant Old English manuscript. The most plausible explanation is not that the relevant changes were incomplete even late in the Old English period but that the traditional language of verse preserved phonological values that had long since changed. There are, however, limits to the conservative power of verse tradition, and some archaic features must be regarded almost certainly as genuinely indicative of the composition of some poems long before the late 10th-century versions in which most verse is recorded (see the discussion in Fulk et al. 2008: clxv–clxvii).
2.2 Morphology Anglian morphological features in verse are much more restrictively distributed than phonological ones: morphological differences are considerably more pronounced than phonological between poems known to have been composed south of the Thames and poems of presumably Anglian provenance. An exception to this generalization is the use of unsyncopated forms of the second and third persons singular of strong verbs (as in bindeþ, WSax. bint ‘binds’), of long-stemmed weak verbs of the first class (as in cy¯ðeð, WSax. cy¯ðð or cy¯ð ‘reveals’), and of preterit/passive participles of longstemmed weak verbs of the first class with stems ending in an oral dental stop (as in læ¯ded, WSax. læ¯dd or læ¯d ‘led’). Unsyncopated forms, alternating with syncopated ones, are to be found even in compositions of West Saxon and Kentish origin (though syncopated forms do not, conversely, appear in poems of Anglian origin). Most other Anglian features are found rarely or not at all in southern compositions, and so they are more likely indications of dialect origins than features of the koine. But because most poetry by far would appear to be Anglian in origin, these features might be said generally to characterize the language of poetry. Some of the most readily recognizable features are these: (1) (2) (3) (4) (5) (6) (7) (8)
consistent use of the verb stem lifi(g)- rather than WSax. libb- ‘live’; first person singular present indicative ending -u/o rather than WSax. -e; preterit of cuman ‘come’ as cwo¯m(-) rather than WSax. co¯m(-); preterit plural se¯gon (WSax. sa¯won) and participle segen (WSax. sewen) to se¯on ‘see’; unstressed preposition fore ‘before’ (WSax. for); preterit plural de¯don ‘did’ (WSax. dydon), sometimes detectable only metrically behind the WSax. orthography; nominative and accusative plural fe¯ondas ‘enemies’ and fre¯ondas ‘friends’ (EWSax. fı¯end, frı¯end); accusative pronouns mec ‘me’, þec ‘you (SG)’, incit ‘you (DUAL)’, u¯sic ‘you (PL)’ (WSax. me¯, þe¯, inc, u¯s).
The tradition-bound language of verse also preserves certain morphological archaisms that are rare in or missing altogether from prose, regardless of dialect. One of these is
25 Old English: Literary language so-called uninflected infinitives after to¯, for example to¯ friclan ‘to request’ (usually to¯ friclanne), though uninflected usage is frequently detectable only on a metrical basis, due to scribal alteration to the more usual, inflected form (Sievers 1885: 255–256, 312, 482). Another is the use of genitive and dative gehwæ¯m ‘each’ in reference to feminine nouns, again sometimes apparent only in scansion, since scribes may substitute analogical LWSax. gehwæ¯re. Limited almost exclusively to the poem Beowulf are i-stem genitive plurals like Deniga ‘Danes’ (beside newer, analogical Dena) and spellings of (-)þe¯o ‘slave’ (mostly in names) without final w.
2.3 Syntax It should be noted that every verse (i.e., half-line) tends to be a syntactically complete constituent (Minkova 2003: 40–41). It is uncommon, for example, for an attributive adjective to modify a noun in the next verse. The syntax of verse is more difficult to analyze than that of prose, in part because the much more extensive use of apposition and the more frequent suppression of pronominal subjects and objects in poetry render ambiguous the relation of many clauses to one another. But this ambiguity is effected chiefly by the nonconformity of verse to normal prose constraints on what sorts of constituents may appear in the crucial second position of clauses. In prose, the second position in principal clauses (after, among many possibilities, a subject, which may be phrasal, or after an adverbial element, such as a prepositional phrase) is normally occupied by a finite verb, while in dependent clauses the second position is normally occupied by any element other than a finite verb, which very commonly appears finally (as in New High German). Thus, for example, þa¯ wæs he¯ a¯risen is independent, ‘then he had arisen’, while þa¯ he¯ a¯risen wæs is dependent, ‘when he had arisen’. Attempts to analyze the syntax of verse according to the same patterns of word order have met with little acceptance (to the arguments of Andrew 1940, 1948, cf., e.g. Mitchell 1985, Vol. II: 88–94). And as the example illustrates, the ambiguity is fed by the homonymy of many adverbs and conjunctions, e.g. þonne ‘then/when’, þæ¯r ‘there/where’, and nu¯ ‘now/now that’. Aside from such ambiguity of clause dependencies, the most striking syntactic feature of verse pertains to the treatment of certain words of variable stress, referred to as particles, which occupy an intermediate prosodic position between fully stressed content words and unstressed clitics. The particles comprise chiefly finite verbs, pronouns, and adverbs of low lexical salience (e.g. demonstrative adverbs, such as þa¯ ‘then’ and þæ¯r ‘there’). In prose, the position of such particles (aside from finite verbs) is relatively unconstrained, while in verse there is a strong tendency for them to cluster at or near the beginning of clauses. Precise formulation of the constraint depends upon both syntactic and prosodic considerations, since particles, in order to remain unstressed, must appear in the first sequence of metrically unstressed syllables in a clause, either before the first fully stressed element (e.g. unstressed adverb þa¯ ‘then’ in þa¯ wæs on burgum ‘then there was in fortresses’, with the first stress on burgum) or immediately after it, when the stressed element is not preceded by any unstressed syllable (e.g. in bugon þa¯ to¯ bence ‘(they) bent then to (the) bench’, i.e., ‘they sat down’, with stress on bugon). Otherwise, poetic meter shows, the particle must be stressed, as in a¯le¯don þa¯ ‘(they) laid then’, in which prior stress falls on the syllable -le¯- as well as on þa¯. This principle is referred to as the Satzpartikelgesetz, or Kuhn’s first law, after its discoverer Hans Kuhn (1933: esp. 8–11). Beowulf is the
389
390
III Old English most conservative poem in a Germanic language in regard to the law, containing fewer than ten exceptions, not all of which are secure. (See Fulk et al. 2008: 324. Kuhn’s laws have provoked considerable controversy. For references, see Momma 1997. Studies that have appeared subsequently include Getty 1997, Orton 1999, Mines 2002, and Suzuki 2002. A particularly perceptive defense of the first law is offered by Donoghue 1997: 71–76.) Even clitics (a category that includes prepositions, pronominal adjectives, coordinating conjunctions, and most prefixes) may bear stress in verse under certain conditions. Thus, a preposition is stressed when it does not stand immediately before its object (e.g. Scedelandum in ‘in Ska˚ne’, Beowulf 19b; see Lapidge 2006) or, occasionally, if its object is a pronoun (e.g. and æfter þon ‘and after that’, Phoenix 238b, with initial stress on the last two words). Likewise, a pronominal adjective is stressed when postposed (e.g. grundwong þone, literally ‘ground-plain that’, Beowulf 2588a). Normally, however, clitics remain unstressed, with the further restriction that the unstressed opening of a clause in verse must not comprise only clitics; it must contain at least one particle. The principle is referred to as the Satzspitzengesetz, or Kuhn’s second law. It is violated more frequently than Kuhn’s first law: thus, Orton (1999: 298) finds some thirteen reliable instances of exceptions in Beowulf, including, for example, þone cwealm gewræc ‘avenged that killing’ (107b), which begins a clause. A small number of syntactic features characteristic of the Anglian dialects are also common in verse. The most familiar of these are the use of the accusative case with the preposition mid ‘with’ (only the dative is thus used in West Saxon) and the masculine gender of sæ¯ ‘sea’ (it is usually feminine in West Saxon). Except in late compositions, the poets aim for economy in the use of grammatical words such as demonstratives, pronouns and conjunctions (especially ond ‘and’), which are deployed much less often than in prose. Verbs of motion are also frequently omitted when an auxiliary will suffice. In view of the opposition between paratactic and hypotactic styles discussed below (Section 3.2), it is worthy of note that the syntax of verse, even disregarding its frequent appositions, can be appreciably complex. An example, (1), is Beowulf 1441b–1454, describing the hero’s arming for combat with Grendel’s mother. (Beowulf is cited from Fulk et al. 2008, except that overpunctuation has been omitted. Other poetic texts are cited from Krapp and Dobbie 1931–1953, except that macrons have been added to indicate vowel quantities.) (1)
Gyrede hine Be¯owulf æ ¯ dum, nalles for ealdre mearn; eorlgew scolde herebyrne hondum gebro¯den, sı¯d ond searofa¯h, sund cunnian, se¯o ðe ba¯ncofan beorgan cu¯þe, þæt him hildegra¯p hreþre ne mihte, eorres inwitfeng, aldre gesceþðan; ac se hwı¯ta helm hafelan werede, se¯ þe meregrundas mengan scolde, se¯can sundgebland since geweorðad, befongen fre¯awra¯snum, swa¯ hine fyrndagum worhte wæ¯pna smið, wundrum te¯ode,
25 Old English: Literary language besette swı¯nlı¯cum, þæt hine syðþan no¯ brond ne beadome¯cas bı¯tan ne meahton. (Beowulf 1441b–1454; Fulk et al. 2008) ‘Beowulf appareled himself with manly costume, cared not at all for his life; his warmail-shirt, interlinked by hand, broad and cunningly decorated, was to make trial of its swimming ability, [that armor] which knew how to defend the bone-chamber [i.e. body] so that no war-grasp, no malicious assault of an angry [enemy], could harm the life in his breast; but the bright helmet protected the head, [the helmet] which was to stir up the lakebed, seek out the intermixed waters, adorned with treasure, enclosed by a curtain of chain-mail, just as a weaponsmith had designed it in days of old, wondrously fashioned it, embellished it with boar-images, so that afterward no swords or battle-blades could penetrate it’. The passage combines typical poetic apposition and asyndetic parataxis with various kinds of syntactic dependencies, including participial phrases, relative clauses, and clauses of result and manner. Many similar passages could be cited.
2.4 Lexis and semantics Even more than its formal features of meter and alliteration, what distinguishes Old English verse from prose is its store of poetic vocabulary. Much of this is poetic in the sense that it is used exclusively, or nearly so, in verse, but some is poetic in the sense that the meanings of a word in verse may differ from those found in prose (see Frank 1986; Griffith 1991; Cronan 2003). Thus, for example, ford in prose has its modern meaning, while in poetry (it appears only in Beowulf) it seems to refer to the sea; and prior to the reign of Knut (1016–1035), eorl in prose referred only to a Scandinavian nobleman, while in verse it meant simply ‘man’ (McKinnell 1975). Poetic simplices are generally assumed to be words that passed out of everyday vocabulary (like erst, ope, and falchion in Modern English), and this assumption is lent support by the fact that some poetic words in Old English have cognates that are exclusively poetic also in cognate traditions, for example þengel ‘prince’ and hæle(ð) ‘hero’ (cf. the cognates, Old Icelandic þengill and halr, both poetic). This analysis is surely correct: most Old English poetry is deeply traditional, aiming to summon up remembrance of a Germanic heroic age, and so antique vocabulary is designed to lend poetic authority to compositions by associating them with ancient traditions (cf., for example, billow and main for ‘wave’ and ‘sea’ in 19th-century poetry). Nouns and adjectives, it has been noted, are the chief varieties of poetic vocabulary; not many verbs or adverbs can be called poetic (Godden 1992: 501). Increasing the store of poetic vocabulary is the extensive use of compound nouns and adjectives. As in the other Germanic languages, compounds are also common in prose, where they generally have a lexicalized, semantically fixed quality, while in verse, compounding is more spontaneous, many poetic compounds appearing as hapax legomena, such as heals-gebedda ‘consort (lit. neck-bedfellow)’ and sadol-beorht ‘saddle-bright’. Compounds of a particularly metaphoric nature are referred to as kennings (though not all kennings are compounds); examples are hron-ra¯d ‘whale-road’, i.e. ‘ocean’, and feorh-hu¯s ‘soul-house’, i.e. ‘body’. In addition to enlarging the fund of poetic diction, compounding serves the purpose of facilitating alliteration. This is evident in, for example, the example (2):
391
392
III Old English (2)
Gamele ne mo¯ston ha¯re heaðorincas, hilde onþe¯on, gif him mo¯dheapum mægen swı¯ðrade. (240b–242; Krapp and Dobbie 1931–53, Vol. 1: 99) ‘The elderly were not permitted, hoary battle-warriors, to serve in combat, if for them, fit in mind, their strength had diminished’.
Here heaðo-rincas ‘battle-warriors’, as an appositive to gamele ‘the elderly’, is not required by syntax or sense, but it serves to provide the alliteration on h that makes possible the off-verse hilde onþe¯on ‘serve in combat’, which is a narrative essential. Likewise, mo¯d-heapum ‘fit in mind’ is not essential information, but it provides the alliteration on m that is required by the off-verse mægen swı¯ðrade ‘their strength had diminished’, demanded by sense and syntax. As a result of the use of such alliterative devices, frequent apposition is a notable feature of verse syntax (see Robinson 1985). The alliterative function of compounding is especially transparent in the interchange of ethnic names like Beorht-Dene ‘Brilliant-Danes’, Ga¯r-Dene ‘Spear-Danes’, Norð-, ¯ ast-, West-Dene ‘North-, South-, East-, West-Danes’, and so forth, which are Su¯ð- E used without distinction of meaning (see Niles 1983: 138–151). Such examples illustrate the nature of formulaic language in Old English verse. That is, whereas formulas in clas¯ o¯´s ‘rosysical epic are fairly rigid, set expressions (like Homeric rhododa´ktulos E fingered Dawn’) designed to fill a particular metrical requirement, Old English formulas are more flexible, being adaptable to various metrical and alliterative needs.
2.5 Pragmatics and rhetoric It seems impossible to prove that there is any difference between the pragmatics of verse and of prose. Yet they may seem different, because representations of speech (especially representations not translated from Latin) are commoner in verse. Due to the formality of poetic discourse, dialogue usually bears less of a resemblance to linguistic interaction than to set speeches. The best poetic source of dialogue is Beowulf, and yet its testimony is impaired by the likelihood that its dialogue is designed, for poetic effect, to reflect a gradual disintegration of communicative ability as the poem progresses (Bjork 1994). In a corpus of about 30,000 poetic lines in a dead language, it is of course difficult to discern any distinctively Anglo-Saxon principles of pragmatics. Nonetheless, Shippey (1993), after examining speeches in Beowulf that illustrate the operation of such pragmatic principles as the Co-operative Principle of Grice (1975), the Politeness Principle of Leech (1983), and the Face Threatening Act studied by Brown and Levinson (1987), argues convincingly for a Conflictive Principle that characterizes Old English heroic speech: In all verbal exchanges, ensure that one’s own worth is stated and acknowledged. If it is acknowledged by hearer, be prepared to acknowledge hearer’s worth. If not, respond with an appropriate degree of reciprocal non-acknowledgement (121).
The pattern is discernible in many speeches in Beowulf, including the hero’s encounters with the shore watch (237–300), with Wulfgar (333–355), and, for the first time, with Hrothgar (407–455).
25 Old English: Literary language Certain rhetorical patterns, more likely of classical than of native inspiration, are common to both prose and verse, for example polysyndeton, as in example (3): (3)
He¯r bið feoh læ¯ne, he¯r bið fre¯ond læ¯ne, he¯r bið mæ¯g læ¯ne. he¯r bið mon læ¯ne (The Wanderer 108–109; Krapp and Dobbie 1931–53, Vol. 3: 137) ‘Here wealth is fleeting, here a friend is fleeting, here a man is fleeting, here a kinsman is fleeting’.
Likewise, it has been maintained that poetry is pervaded by elaborate patterns of ringcomposition, whereby paired compositional elements successively enclose similarly paired elements, in a pattern like the layers of a halved onion (see, e.g., Niles 1983: 157–162), a necessarily literate pattern, and one previously claimed for many classical compositions, including Homeric and Virgilian epic. A few rhetorical patterns are peculiar to verse, or at least are commoner there, and thus they are likely to reflect native Germanic compositional habits. The classical poetic style is characterized by a high proportion of enjambed lines, in which there is no major syntactic division between one line and the next, and clauses tend to begin directly after the mid-line caesura, as does, for example, the sentence beginning at Exodus 240b, quoted above (for references, see Calder 1979: 37–39). The pattern contrasts markedly with early Scandinavian verse, which is strophic, and with late and relatively prosaic (i.e., non-classical) Old English verse, such as Instructions for Christians and Judgment Day II, which, with their predominantly end-stopped lines, are rhetorically more monotonous. Notable also in verse of the classical style is the pattern of closing a passage with an off-verse that is a complete sentence, providing a kind of aural punctuation. For example, the praise of God that closes Andreas is capped by a succinct assessment, the last verse in the poem (1722b): Þæt is æðele cyning ‘That is a noble king’. Cf. Þæt wæs go¯d cyning (Beowulf 11b, 2390b), serving a similar function (see Fulk 1996: 77–78). Also, contrast is a controlling principle in Anglo-Saxon thought. In poetry it produces a common rhetorical structure wherein a negative proposition (often introduced by ne ‘not’) is invoked in order to affirm its immediately following positive opposite (introduced by ac ‘but’), for example in The Battle of Maldon 81–83, where reference is made to Ælfhere and Maccus, defenders of the ford, (4) (4)
fle¯am gewyrcan þa¯ noldon æt þa¯m forda wið ða¯ fy¯nd weredon ac hı¯ fæstlı¯ce, æ ¯ pna wealdan mo¯ston. þa¯ hwı¯le þe hı¯ w (The Battle of Maldon 81–83; Krapp and Dobbie 1931–53, Vol. 6: 9) ‘who would not at the ford take flight, but they firmly made defense against the enemy, the while that they could wield weapons’.
Not infrequently in verse (though rarely in prose), the positive opposite is omitted from the comparison, resulting in a kind of understatement (litotes or meiosis), as when the poet of Beowulf expresses the pleasure of humans at the death of Grendel: No¯ his lı¯fgeda¯l / sa¯rlic þu¯hte secga æ¯negum ‘His parting from life did not seem distressing to anyone’ (841b–842). But understatement may take many forms, as when, in Beowulf, it is
393
394
III Old English said of a sea-beast killed by an arrow, he¯ on holme wæs / sundes þe¯ sæ¯nra ‘in the sea it was slower at swimming’ (1435b–1436a; see Bracher 1937).
3 The language of literary prose The corpus of non-poetic Old English is varied in its textual types, including wisdom and travel literature, penitentials, medical books, chronicles, laws, homilies and works of theology, letters, and many shorter medical recipes, prognostics, liturgical and penitential texts, inscriptions, records, and charters, as well as countless glosses. In literary studies, it is no longer the norm to distinguish literary from other types of texts, but for linguistic purposes the distinction is a useful one, since the language of texts composed as continuous prose differs markedly from, for example, that of books of law, penance, and medicine, which were composed for reference rather than sustained reading (and hence are often cryptic in their terseness), and from that of interlinear glosses, which were designed to provide guidance as to sense rather than to be read through as full-fledged translations, even when every word of a text was glossed. For present purposes, then, the discussion will be limited to literary prose, which comprises chiefly homilies, letters, and works translated from Latin literary prose, such as the Old English Apollonius of Tyre, Bede’s ecclesiastical history, travel literature, and the works of the Alfredian program of translation (Orosius, Boethius, Gregory’s Cura pastoralis, Augustine’s Soliloquia, and the prose Psalms). The Anglo-Saxon Chronicle may also be considered literary. Once literary prose is thus defined, it can be seen that no prose of any substance survives in a dialect other than West Saxon. Literary texts nonetheless conform to one of three West Saxon types: (1) those in Early West Saxon (the Alfredian texts just listed); (2) those in standard Late West Saxon, which dialect is defined by its conformity to the standards of Æthelwold and his student Ælfric; (3) other Late West Saxon texts, which may simply lack many of the defining features of the Æthelwoldian dialect (as do, for example, most copies of works by the homilist and jurist Wulfstan), but which may show a remarkable admixture of such seemingly Anglian and Kentish features as characterize most of the poetry. Texts of the last sort include nearly all the anonymous homilies (although the nonstandard features vary considerably in variety and incidence in this group). Anglian features are also particularly pronounced in Solomon and Saturn I & II, Scriftboc, the Herbarium and Medicina de quadrupedibus, Lacnunga, Bald’s Leechbook, the translation of Boniface’s letter to Eadburga, Alexander’s Letter to Aristotle, and The Marvels of the East. Less frequent Anglian features are evident in the Worcester and Peterborough Chronicles, and in the translation of capitula 1–16 of Alcuin’s De virtutibus et vitiis in London, British Library, Cotton Vespasian D. xiv. Mercian features are especially frequent in some earlier texts: Bishop Wærferth’s translation of Gregory the Great’s Dialogi, the Old English Bede, and the various fragments of the Old English Martyrology (see Fulk 2008). The reasons for this admixture of nonstandard features is disputed: it may be due to original composition wholly or partly in another dialect, with subsequent “translation” into West Saxon (as is certainly the case with, for example, the Old English Bede: see J. Campbell 1951), but it may be due to regional or social variation within West Saxon, to competition among different artificial, literary standards or focused varieties, to register, to scribal idiosyncrasies, or to a combination of such factors.
25 Old English: Literary language
3.1 Phonology, lexis, and rhetoric The homilies of Ælfric (Clemoes and Godden [eds.] 1997–2000; Pope [ed.] 1967–68) make up a large part of the prose corpus, and a significant number of these are composed in an alliterative prose that bears a superficial resemblance to verse. It is divisible into lines, but it lacks a definite metrical structure, its vocabulary is rarely poetic, apposition plays no significant role, and the placement and number of alliterative staves are freer than in verse. Ælfric himself seems not to have regarded this as poetry, as he refers at one place to composing on u¯re wı¯san ‘after our manner’, as opposed to on le¯oðwı¯son ‘in verse’. (The two phrases are used, respectively, in Ælfric’s letter for Sigeweard and in the ‘Excusatio dictantis’ appended to his homily on the deposition of St. Martin, no. XXXIV in the second series of Catholic Homilies.) The less numerous homilies of Ælfric’s contemporary Wulfstan are also marked by a heightened rhetorical style. They make frequent use of pleonastic binomials like gesæ¯lig and e¯adig ‘prosperous and fortunate’ and unwı¯sdom and swicdom ‘foolishness and error’, which sometimes alliterate (e.g. habban and healdan ‘to have and to hold’), though they more frequently rhyme (e.g. stalu and cwalu ‘theft and killing’ and sacu and clacu ‘strife and injury’). Wulfstan also favors certain kinds of Latinate parallelisms (see Bethurum 1957: 91) and frequent intensifiers, such as æ¯fre ‘ever’, ealles to¯ swı¯ðe ‘all too much’, and oft and gelo¯me ‘again and again’. The appearance of Anglian vocabulary in anonymous prose (e.g. nymðe ‘unless’ and oferhygd ‘pride’) possibly indicates the Anglian origins of the text. When, on the other hand, Anglian words and forms are to be found in the West Saxon writings of Ælfric and Wulfstan, they are presumably a mark of elevated, quasi-poetic diction. Thus, for example, Ælfric uses the Anglian verb lifigan in ðone lifigende God ‘the living God’ and synonymous expressions, whereas everywhere else he uses the West Saxon verb libban. He also uses the poetic words metod (in reference to God) and heolstor ‘darkness’, and some others (see Godden 1980: 217–219; Frank 1994). Conversely, just as there was a body of exclusively poetic diction, some vocabulary was plainly regarded as prosaic (most of it, probably, neologistic) and unsuited to verse. Examples are ce¯pan ‘seize’, cnapa ‘child’, fultum ‘assistance’, macian ‘make’, namian ‘name’, and wı¯fmann ‘woman’. The occurrence of prosaic words in verse (e.g. hopian ‘hope’ in Judith) is now generally taken to be a sign of a poem’s late composition (see Stanley 1971).
3.2 Syntax The syntax of texts that are (presumably) not translated from Latin can be remarkably varied. The style of the Anglo-Saxon Chronicle generally gives the impression of simplicity, rarely deviating from a pattern of short paratactic clauses, most commonly conjoined by and. This style is no doubt largely dictated by the content: the annals record events in sequence, with little or no analysis or comment. By contrast, the syntax of King Alfred’s more literary epistle prefaced to his translation of Gregory the Great’s Cura pastoralis can be quite complex, with multiple varieties of hypotaxis, as in this much-remarked example, (5): (5)
Forðy¯ me¯ ðyncð betre, gif ¯ıow swæ¯ ðyncð, ðæt we¯ e¯ac sumæ be¯c, ða¯ ðe nı¯edbeðearfosta sı¯en eallum monnum to¯ wiotonne, ðæt we¯ ða¯ on ðæt geðı¯ode wenden ðe we¯ ealle gecna¯wan mægen, and gedo¯n swæ¯ we¯ swı¯ðe e¯aðe magon mid
395
396
III Old English Godes fultume, gif we¯ ða¯ stilnesse habbað, ðæt eall sı¯o gioguð ðe nu¯ is on Angelcynne frı¯ora monna, ða¯ra ðe ða¯ spe¯da hæbben ðæt hı¯e ðæ¯m befe¯olan mægen, sı¯en to¯ liornunga oðfæste, ða¯ hwı¯le ðe hı¯e to¯ na¯nre o¯ðerre note ne mægen, oð ðone first ðe hı¯e wel cunnen Englisc gewrit a¯ræ¯dan; læ¯re mon siððan furður on Læ¯dengeðı¯ode ða¯ ðe mon furðor læ¯ran wille and to¯ hı¯eran ha¯de do¯n wille. (King Alfred’s translation of Gregory’s Pastoral Care 6–15; Sweet [ed.] 1871: 6) ‘Therefore it seems to me better, if it seems so to you, that we also translate certain books – those that are most necessary for all people to know – into that language that we can all understand, and arrange (as we very easily can with God’s help, if we have the [necessary] respite from war) that all the sons now in England of those free men who have the wherewithal that they can be devoted to it, be committed to schooling, for the period that they cannot [be put] to other employment, until the time that they can well read writing in English; afterward, let those be taught further in the Latin language whom one would like to teach further and place in a higher [i.e. ecclesiastical] order’. The subject of the sentence is an elaborate nominal clause beginning with the first ðæt, in which various subordinate clauses are embedded, including a number of relative clauses, adverbial clauses of time and manner, a conditional clause, and a clausal complement dependent on the notion of sufficiency that is implied but unexpressed in spe¯da ‘wherewithal’. On the whole, Old English lends itself better to this degree of complexity than does Modern English, as represented by the translation. For example, whereas the construction ‘sons […] be committed to schooling’ in the translation is awkward because of the length of the intervening material containing modifiers, relative clause, and clausal complement, there is no ambiguity of structure in the Old English. The only noticeable sign of possible awkwardness in the Old English is the pleonasm (eliminated in the translation above) of ðæt we¯ e¯ac sumæ be¯c […] ðæt we¯ ða¯ on ðæt geðı¯ode wenden, literally ‘that we also certain books […] that we them into that language translate’, occasioned by the intervening relative clause. But the seeming awkwardness is perhaps due only to the avoidance of such pleonasms in Modern English. Old English was plainly less rigidly Latinate in its syntactic logic (e.g. requiring multiple negation in negative clauses containing indefinite elements), and the structure of the Old English appears both idiomatic and transparent. The native and natural quality of such syntax is suggested by the fact that literature translated from Latin does not normally reach such a degree of complexity, and, in general, the less slavishly dependent a translation is (like, e.g., Alfred’s translation of Boethius), the likelier the syntax is to be complex. Conversely, translations often evince syntactic structures that are more likely to reflect Latinate syntax than native idiom. Particularly noteworthy are dative constructions resembling the Latin ablative absolute, for example swa¯pendum windum ‘when the winds were blowing’ (Bede 3, 14.202.14, rendering ventis ferentibus), though these do not always translate Latin ablatives (see Mitchell 1985, Vol. II: 914–937, with references).
4 Summary The language of Old English poetry is understandably more formal and artificial than that of prose, especially in its lexis and syntax, though graphemic features corresponding
25 Old English: Literary language to Anglian phonological traits also attest to the elevated register of the genre, given the historical prestige of the Anglian communities. Certain pragmatic and rhetorical features are discernible more plainly in verse than in prose, probably because of the dialogic and heroic mode of much poetry. In regard to prose it is more difficult to distinguish native features of literary language because most prose is translated from Latin. Even though nearly all prose in the sense employed here is preserved in the Early or Late West Saxon dialect, it is far from uniform in nature, much of it showing an admixture of non-West Saxon features that correlate most probably to both register and textual history (i.e., derivation from Anglian originals), though regional variation may also have played a role. Much of this variation is plainly intentional, and the crafted styles of such writers as Ælfric and Wulfstan attest to the value attached to varieties of literary language neither wholly poetic nor prosaic.
5 References Andrew, S. O. 1940. Syntax and Style in Old English. Cambridge: Cambridge University Press. Andrew, S. O. 1948. Postscript on Beowulf. Cambridge: Cambridge University Press. Bethurum, Dorothy (ed.). 1957. The Homilies of Wulfstan. Oxford: Oxford University Press. Bjork, Robert E. 1994. Speech as gift in Beowulf. Speculum 69: 993–1022. Bracher, Frederick. 1937. Understatement in Old English poetry. PMLA 52: 915–934. Brown, Penelope, and Stephen C. Levinson. 1987. Politeness: Some Universals in Language Use. 2nd edn. Cambridge: Cambridge University Press. Calder, Daniel G. 1979. The study of style in Old English poetry: A historical introduction. In: Daniel G. Calder (ed.), Old English Poetry: Essays on Style, 1–65. Berkeley: University of California Press. Campbell, A. 1969 [1959]. Old English Grammar. Reprinted from corrected sheets of the 1st edn. Oxford: Clarendon. Campbell, Jackson J. 1951. The dialect vocabulary of the Old English Bede. Journal of English and Germanic Philology 50: 349–372. Clemoes, Peter and Malcolm Godden (eds.). 1979–2000. Ælfric’s Catholic Homilies. 3 vols. (Early English Text Society, S. S. 5, 17, 18.) Oxford: Oxford University Press. Cronan, Dennis. 2003. Poetic meanings in the Old English poetic vocabulary. English Studies 84: 397–425. Donoghue, Daniel. 1997. Language matters. In: Katherine O’Brien O’Keeffe (ed.), Reading Old English Texts, 59–78. Cambridge: Cambridge University Press. Frank, Roberta. 1986. “Mere” and “sund”: Two sea-changes in Beowulf. In: Phyllis Rugg Brown, Georgia Ronan Crampton, and Fred Robinson (eds.), Modes of Interpretation in Old English Literature: Essays in Honour of Stanley B. Greenfield, 153–172. Toronto: University of Toronto Press. Frank, Roberta. 1994. Poetic words in Late Old English prose. In: Malcolm Godden, Douglas Gray and Terry Hoad et al. (eds.), From Anglo-Saxon to Early Middle English: Studies Presented to E. G. Stanley, 87–107. Oxford: Oxford University Press. Fulk, R. D. 1992. A History of Old English Meter. Philadelphia: University of Pennsylvania Press. Fulk, R. D. 1996. Rhetoric, form, and linguistic structure in early Germanic verse: Toward a synthesis. Interdisciplinary Journal for Germanic Linguistics and Semiotic Analysis 1: 63–88. Fulk, R. D. 2008. Anglian dialect features in Old English anonymous homiletic literature: A survey, with preliminary findings. In: Susan Fitzmaurice and Donka Minkova (eds.), Empirical and Analytical Advances in the Study of English Language Change, 81–100. Berlin/New York: Mouton de Gruyter.
397
398
III Old English Fulk, R. D., Robert E. Bjork, and John D. Niles (eds.). 2008. Klaeber’s “Beowulf” and the Fight at Finnsburg. 4th edn. Toronto: University of Toronto Press. Getty, Michael. 1997. Was finite verb placement in Germanic prosodically conditioned?: Evidence from Beowulf and Heliand. Journal of English and Germanic Philology 96: 155–181. Godden, Malcolm. 1980. Ælfric’s changing vocabulary. English Studies 61: 206–223. Godden, Malcolm. 1992. Literary language. In: Richard M. Hogg (ed.), The Cambridge History of the English Language. Vol. I. The Beginnings to 1066, 490–535. Cambridge: Cambridge University Press. Grice, H. P. 1975. Logic and conversation. In: Peter Cole and Jerry L. Morgan (eds.), Speech Acts, 41–58. New York: Academic Press. Griffith, M. S. 1991. Poetic language and the Paris Psalter: The decay of the Old English tradition. Anglo-Saxon England 20: 167–186. Krapp, George Philip and Eliott Van Kirk Dobbie (eds.). 1931–53. The Anglo-Saxon Poetic Records: A Collective Edition. 6 vols. New York: Columbia University Press. Kuhn, Hans. 1933. Zur Wortstellung und -betonung im Altgermanischen. Beitra¨ge zur Geschichte der deutschen Sprache und Literatur 57: 1–101. Lapidge, Michael. 2006. An aspect of Old English poetic diction: The postpositioning of prepositions. In: John Walmsley (ed.), Inside Old English: Essays in Honour of Bruce Mitchell, 153– 180. Oxford: Blackwell. Leech, Geoffrey. 1983. The Principles of Pragmatics. London: Longman. McKinnell, John. 1975. The date of The Battle of Maldon. Medium Ævum 44: 121–136. Mines, Rachel. 2002. An examination of Kuhn’s second law and its validity as a metrical-syntactical rule. Studies in Philology 99: 337–355. Minkova, Donka. 2003. Alliteration and Sound Change in Early English. Cambridge: Cambridge University Press. Mitchell, Bruce. 1985. Old English Syntax. 2 vols. Oxford: Clarendon. Momma, H. 1997. The Composition of Old English Poetry. Cambridge: Cambridge University Press. Niles, John D. 1983. Beowulf: The Poem and Its Tradition. Cambridge, MA: Harvard University Press. Orton, Peter. 1999. Anglo-Saxon attitudes to Kuhn’s laws. Review of English Studies N. S. 50: 287–303. Pope, John C. (ed.). 1967–68. Homilies of Ælfric: A Supplementary Collection. 2 vols. (Early English Text Society, O. S. 259–60.) Oxford: Oxford University Press. Pope, John C. (ed.). 2001. Eight Old English Poems. 3rd edn. rev. by R. D. Fulk. New York: W. W. Norton. Robinson, Fred C. 1985. Beowulf and the Appositive Style. Knoxville: University of Tennessee Press. Shippey, T. A. 1993. Principles of conversation in Beowulfian speech. In: John M. Sinclair, Michael Hoey, Gwyneth Fox, and Malcolm Coulthard (eds.), Techniques of Description: Spoken and Written Discourse: A Festschrift for Malcolm Coulthard, 109–126. London: Routledge. Sievers, Eduard. 1885. Zur Rhythmik des germanischen Alliterationsverses. Beitra¨ge zur Geschichte der deutschen Sprache und Literatur 10: 209–314, 451–545. Sievers, Eduard. 1893. Altgermanische Metrik. Halle: Niemeyer. Sisam, Kenneth. 1953. Studies in the History of Old English Literature. Oxford: Clarendon. Stanley, E. G. 1971. The prosaic vocabulary of Old English verse. Neuphilologische Mitteilungen 72: 385–418. Suzuki, Yasuko. 2002. The prosody and syntax of light elements in West Germanic alliterative verse: With special reference to Beowulf. In: Irmengard Rauch and Gerald F. Carr (eds.), New Insights in Germanic Linguistics II, 225–250. New York: Peter Lang. Sweet, Henry (ed.). 1871. King Alfred’s West-Saxon Version of Gregory’s Pastoral Care. 2 vols. (Early English Text Society, O. S. 45, 50.) Oxford: Oxford University Press.
Robert D. Fulk, Bloomington (USA)
IV Middle English 26 Middle English: Phonology 1 2 3 4 5 6
Introduction Word stress Rhythm Segmental phonology Summary References
Abstract This chapter describes segmental and suprasegmental developments in Middle English phonology. It identifies major historical trends as well as the causalities behind them. Its focus on explanation also motivates a departure from the traditional order of presentation. Taking into account that rhythmic patterns are historically more stable than the properties of segments and impose strong constraints on their transmission, this chapter discusses Middle English word stress first, and explains how the integration of French loan words affected the system behind it. Next the rhythmic patterns are described which emerged from the expression of lexical stress and which constituted the environment for the phonetic realization of Middle English speech sounds. The final section describes developments in segmental phonology, both in unstressed and in stressed environments. An attempt is made to show how most of them can be understood as adaptive responses to rhythmic constraints on the phonetic realization of segments.
1 Introduction Any historical survey of a language is prone to essentialist misinterpretation. One needs to remember that no such thing as a “Middle English phonology” ever existed. Accounts like this generalize over large populations of individual speakers’ phonological competences and their constituents, which all had particular distributions, specific functions in various subgroups of English society, and specific histories. While aspects of regional variation in Middle English (ME) phonology are covered in Williamson (Chapter 31), this summary attempts to distill general trends from sociohistorical diversity and to expose potential causalities in the way in which they unfolded.
2 Word stress Word stress is a good starting point for a survey like this, because many aspects of Middle English segmental and syllabic phonology relate to it. It also represents an area where Middle English deserves the “Middle” in its name, because the period saw a transition between two different ways of stress assignment: at its beginning the position of primary stress reflected the so-called Germanic Stress Rule (GSR), which assigned stress to the first syllable of lexical roots. At the end of the period, primary stress Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 399–414
400
IV Middle English placement reflected different principles, and seems to have depended more strongly on factors like the weight of syllables or their distance from the end of a word. It is therefore often said that Middle English replaced the GSR with a Romance Stress Rule (RSR) (e.g. Lass 1992: 83–90), which it imported together with a large number of French and Latin loans in the wake of the Norman Conquest. By the GSR, stress always fell on the first syllable of a lexical root and stayed there even if inflectional endings, derivational suffixes or prefixes were added. Thus, fā der ‘father’ was stressed ˈfā der, and the stress on ˈfā - is preserved also in inflected forms like ˈfā deres, or derivations like ˈfā der-hō d (N), or ˈfā der-lē s (ADJ). Likewise, the stress on the root syllables in ˈsai-en ‘say’, or ˈsink-en ‘sink’, was retained in prefixed derivations such as forˈsaien ‘renounce’, or forˈsinken, ‘fall into ruins’. In combinations of two lexical roots (i.e. in compounds, or combinations with “root prefixes” such as ˈwiðer- ‘against’, or fore- ‘before’), primary stress fell on the first of the two roots, while the second one retained secondary stress, as in ˈheven-ˌrī che ‘heavenly kingdom’, ˈdē vel-ˌwerk ‘devil’s work’, or ˈwiðer-ˌsaca ‘archenemy’. However, the morphological grounding of the GSR may not have been fully transparent to early Middle English speakers anymore, because some originally derived words had become opaque and were memorized as simple lexical units. Thus they carried main stress on syllables that were no longer perceived as the first ones of their roots. Examples are aˈbū te(n) ‘about’, ʒeˈnō h ‘enough’, aˈʒeines ‘against’, aˈwei ‘away’, biˈleafe ‘faith’, or biˈhā lden ‘behold’. Although Middle English speakers were no longer aware that also in these words stress had once indicated the left root boundary, they would nevertheless have regarded them as perfectly native. This also suggests that the GSR was not an active production rule in Middle English phonology anymore: adult Middle English speakers are unlikely to have invoked it for assigning stress to words after they retrieved them from their mental dictionaries. Instead, they are more likely to have memorized words that were fully specified with regard to their stress patterns, so that the GSR would primarily be used in perception, particularly for identifying violations of the rule in foreign words. This was the system of stress into which French words came to be borrowed. The principles on which they had received their stress in French were different from the Germanic ones. They operated from right to left and assigned stress on the basis of syllable weight. French syllable types fell into two categories. Syllables that ended in a short vowel (V) counted as “light” or “weak” (σw), and syllables that ended in a long vowel (VV), a vowel and a consonant (VC), or more counted as “heavy” or “strong” (σs). If the last syllable of a Medieval French word was heavy, or strong (σs), it received primary stress; if it was light, or weak (σw), stress fell on the penultimate syllable. Thus, Medieval French loans had stress patterns like those shown in (1): (1)
a. (x)ˈσs: ciˈtee ‘city’, deˈgree, diˈvers, ‘diverse’, adversiˈtee ‘adversity’, charitˈee ‘charity’, condiciˈoun ‘condition’, juggeˈment ‘judgement’, carpenˈter b. (x)ˈσσw: ˈdette ‘debt’, engenˈdrure ‘procreation’, pilgriˈmage, couˈrage, maˈtere ‘matter’
Many of these patterns were clearly un-English and were adjusted as the French loans came to be fully integrated into the Middle English lexicon. However, the new loans
26 Middle English: Phonology were not simply subjected to the morphologically grounded Germanic principles of stress assignment. After all, the morphological structures of French words would have made little sense from the point of view of native English morphology, so that a rule which relied on the identification of root morphemes would have been difficult to apply. Thus, French loans did not simply have the GSR applied to them. Instead, they were made to resemble the surface stress contours which the GSR typically produced. On the level of surface phonetics, of course, the morphological composition of a word does not matter as much as its sound shape, the weight of its syllables, and their relative prominence. Therefore, for the purpose of assimilating French loans to native-like stress patterns, those native patterns were themselves interpreted as relations between stress and syllable weight rather than as relations between stress and morphological structure. In this sense Romance principles of weight-based stress assignment came to replace the Germanic way of assigning word stress on morphological grounds. In practice, this worked along the lines set out in Sections 2.1–2.4.
2.1 Monosyllabic words These were not a problem. They had only a single syllable on which stress could fall. However, some French monosyllables, such as de´ ‘die’, could be light (CV), whereas native English monosyllables could not. In such cases the French vowels were reinterpreted as long – e.g. as dē – and were thereby made to fit into the native system.
2.2 Disyllabic words Among English disyllables, there were two types: those with light final syllables (i.e. σσw), and those with heavy ones (σσs). Since Middle English behaved like many of the world’s languages in that it disregarded word-final consonants in the calculation of syllable weight, not only syllables that ended in V counted as light, but also those that ended in VC. Syllables with VV, VCC or more segments in their rhymes were of course heavy. Now, ME disyllables with light final syllables (i.e. σσw) were practically always stressed on the penult (e.g. ˈdrihten ‘lord’, ˈheorte ‘heart’, ˈdohter ‘daughter’, or ˈmaniʒ ‘many’), whereas syllables of the type σσs had variable stress patterns: many of them were monosyllabic roots preceded by unstressed prefixes and would therefore be stressed on their final syllables (e.g. aˈwei, ʒeˈno¯h, ʒeˈsū nd ‘healthy’, ʒeˈwiht ‘weight’, to-ˈdei ‘today’). There were also many σσs disllyables, however, which were stressed on the first syllable (e.g. ˈlauerd ‘lord’, ˈakern ‘acorn’, ˈhǣ lend ‘saviour’, verbal 2P SG like ˈluvest ‘you love’, and superlatives like ˈfairest). In short, native σσw disyllables were practically always ˈσσw, while σσs could be either ˈσσs or σˈσs. For French loans this meant that disyllabic words with light final syllables (like ˈdette) fitted the native ˈσσw pattern with no adjustment at all, while for words like ciˈtee, deˈgree, or diˈvers, a choice had to be made, since there were two native models which they could follow. The choice often seems to have been a matter of chance, and there is evidence that French σσs loans could have occurred with either of their two syllables stressed for some time. The example most often given to illustrate the high level which stress variability of this kind reached is Chaucer’s famous line in (2):
401
402
IV Middle English (2)
in ˈdivers ˈarts and ˈin diˈverse fiˈgures (c.1390 Chaucer, “Friar’s Tale”, Canterbury Tales [CT] D1460)
But other stress doublets of the same type are easy to find, such as (3): (3)
a. In ˈwhich he ˈal the ˈnoble ˈcitee ˈseigh (c.1390 Chaucer, “Knight’s Tale”, CT A208). b. And ˈmak(e) a ˈwerr(e) so ˈsharp on ˈthis ciˈte (c.1390 Chaucer, “Knight’s Tale”, CT A429)
Interestingly, the variability of stress in French σσs disyllables seems to have affected native words as well. Thus, in Chaucer we find both ˈhousbond (“Wife of Bath’s Tale Prologue”, CT D19) and husˈbond (“Wife of Bath’s Prologue”, CT D22). It took considerable time until such ambiguities were eventually resolved by convention, and they were not settled completely systematically. Possibly reflecting the greater frequency of prefixed verbs against prefixed nouns in the native English vocabulary, Romance σˈσs verbs seem to have retained final stress more often than Romance σˈσs nouns. This may explain the contrasting stress patterns in pairs like Modern English ˈprogress (N) vs. proˈgress (V). Yet even today pronunciations such as reˈsearch and ˈresearch (both N) or acˈcess and ˈaccess (both V) are viable alternatives.
2.3 Trisyllabic words Among trisyllabic words of English origin two different stress patterns occurred, namely ˈσσσ and σˈσσ. Their distribution depended on the weight of the medial syllable. On the one hand, there were word forms with light medial syllables, (i.e. σσwσ). Since there were few prefixed word forms (e.g. iˈcleped ‘called’, forˈʒeven ‘forgive’) among them, most of them were stressed on the first syllable (e.g. ˈsō ðliche ‘truly’, ˈluuede(n) ‘(they) loved’, ˈlī come ‘body’ ˈheouene(s) ‘heaven(’s)’). On the other hand, there were also many trisyllabic forms with heavy medial syllables (i.e. σσsσ), and among them no stress pattern had a clear majority. Many of them were prefixed forms and carried stress on the heavy medial syllable, such as alˈmī hti ‘almighty’, aˈbū te ‘about’, or forˈbī sne ‘example’. Many others, however, did not. They included gerundial or participial forms like ˈbletsunge ‘blessing’, ˈlē asunge ‘lie’, ˈhælende ‘saviour, healer’, or ˈcwæðende ‘saying’, and they all had stress on their first syllables. Thus, while most native σσwσ forms were ˈσσwσ, native σσsσ forms could be either ˈσσsσ or σˌσsσ. French trisyllabic loans, on the other hand, usually had either of the following two stress patterns. If their final syllable was heavy (σσσs), it would be stressed, as in chariˈtee or juggeˈment. If their final syllable was light (σσσw), however, stress would fall on the penult, as in serˈvise or enˈdure. Although it appears at first sight to be less compatible with native English patterns, the first of these two types (i.e. σσˈσs) was easier to assimilate. This is because in French loans of the σσˈσs type, the first syllable would have been perceived as more prominent than the medial one and would often carry secondary stress when uttered. Since the relative prominence of two stressed syllables that are separated by unstressed material is a matter of degree, it would easily have gone unnoticed if realizations like ˌchariˈtee, or ˌjuggeˈment were rather interpreted as ˈchariˌtee or ˈjuggeˌment. Therefore, they soon fell in line with native words of the
26 Middle English: Phonology types ˈheouene or ˈcwæðende. With French loans of the type σˈσσw (i.e the type serˈvise/ enˈduren), on the other hand, speakers faced a similar choice as with disyllables of the type ciˈtee/deˈgree. Just like in the corresponding disyllables, the eventual development of naˈture/enˈdure-type trisyllables was not fully predictable, although once again more nouns (Modern English ˈservice) than verbs (Modern English enˈdure) wound up with stress fixed on the first syllable.
2.4 Words with more than three syllables Many native words in this group were compounds or derivatives (e.g. ˈærceˌbiscop, ‘archbishop’, ˈeuelˌnesse ‘evilness’). They usually had primary stress on their first syllables and secondary stress on their penults, and provided natural models for French loans of the type ˌpilgriˈmage. Such loans could easily be adjusted to native patterns through a re-interpretation of relative stress prominence. Thus, ˌpilgriˈmage became ˈpilgriˌmage. Likewise, words such as adˌversiˈtee, or conˌdiciˈoun were first reinterpreted as adˈversiˌtee, or conˈdiciˌoun, lost their secondary stress, and were thereby assimilated to native word forms like forˈʒifnesse ‘forgiveness’. In sum, foreign words of Romance provenance came to assume stress contours that were just like the native patterns which had been created on Germanic principles. At the same time, the growing number of Romance loans reduced the transparency of these principles: since words like adversitee, engendrure, or condicioun were morphologically opaque in Middle English, the notion that they should be stressed on the first root syllable would not have been helpful. Therefore, even as Romance loan words were given English stress contours, the English criteria for assigning word stress were gradually superseded by Romance ones (cf. Minkova 2006: 114–121), and ever since the end of the ME period the better generalizations about primary stress in English words have been derivable by parsing them from right to left and taking syllable weight into account. Nevertheless, the Germanic preference for root initial stress has remained inferable from the large number of English di- and trisyllables with initial stress. Although the criteria for English stress placement changed during the ME period, the principle that English word stress was lexically “fixed” did not. Except for the phase during which stress doublets like diˈvers : ˈdivers, naˈture : ˈnature occured, English words have always had the same syllables stressed, irrespective of the context in which they were uttered. This had been true in Old English, where words like ˈfæder, ˈbisceop, ˈǣ rende, or ˈyfelnesse never showed up as *fæˈder, *bisˈceop, *ǣ renˈde, or *yˈfelnesse, and it is still true in Modern English. In this respect English has resisted Romance influence and has remained different from languages like Italian or French, where one gets ˈmango ‘I eat’ vs. manˈgiamo ‘we eat’, or French je ˈmange ‘I eat’ vs. vous manˈgez ‘you-PL eat’. The principle that English word stress has remained invariable had important consequences for phonotactic shapes of ME words, and on the sounds that made them up.
3 Rhythm Before looking at ME speech sounds and their histories, it is useful to consider the rules that governed ME utterance rhythm. Strongly dependent on word stress, the rhythmic
403
404
IV Middle English configurations in which ME sounds were uttered, exerted a strong influence on their phonetic realizations and consequently on their diachronic development. As in other Germanic languages, the rhythm of ME speech was based on stress feet (see e.g. Dresher and Lahiri 1991), i.e. the alternation of prominent and less prominent syllables. ME feet began with a stressed, or strong syllable (s), and included all the unstressed, or weak ones (w) that followed it before the next stress occurred. Utterance rhythm was closely related to word stress. The stressed syllables of nouns, verbs, and adjectives (i.e. major lexical categories) would usually also be rhythmically prominent, while grammatical morphemes such as prefixes, suffixes, prepositions, pronouns, etc. often showed up rhythmically weak. Although utterance rhythm was strongly constrained by lexical word stress, it did not follow from it mechanically. Instead, rhythmic patterns resulted from post-lexical interactions among the word forms on which they were implemented, and reflected a universal preference for the regular alternation of prominent and less prominent syllables. Prominence is a relative matter: a prominent syllable will be more reliably perceived as prominent if its neighbors are weaker, and vice versa. Since sequences of equally prominent syllables are perceptually confusing and generally dispreferred (cf. Dziubalska-Kołaczyk 2002), they were also avoided in Middle English. For instance, in a sequence of three or more weak syllables flanked by strong ones (i.e. ˈswwwˈs), one of the syllables in the middle would often acquire secondary prominence and become the head of an utterance foot, as the first syllable of -nesse in (4): (4)
Þe | ˈchildes | ˈmichel |ˌnesse | ˈsheude þe | ˈengel on the child’s greatness–ACC showed the angel–NOM by | ˈfuwer | ˈþingen (c.1225 Trinity Homilies, XXIII: 135) four things
or the preposition in in (5): (5)
| ˈsuffre þu | ˈnawt þe | ˈfeond þt ne | ˈlead(e) us | ˈallung(e) suffer you not the devil that NEG lead us ever |ˌin to | ˈfondung(e) (c.1230 Ancrene Wisse 116) in to temptation. ne
NEG
The ability of weak syllables to be raised also depended on their weight, and heavy syllables were more likely to acquire rhythmic prominence than light ones, while ultralight weak syllables, which ended in a short vowel (as in inflectional endings), never did. Just as weak syllables could acquire prominence in the neighborhood of even weaker ones, lexical monosyllables could be rhythmically weakened in the context of stronger ones. Thus, right could both figure as a foothead, as in (6): (6)
And ˈright as ˈfrely ˈas he ˈsente hir ˈme (c.1390 Chaucer, “Franklin’s Tale”, CT F896)
or be demoted in the context of stronger neighbors, as in (7): (7)
For, ˈquyk or ˈdeed, right ˈthere ye ˈshal me ˈfynde. (c.1390 Chaucer, “Franklin’s Tale”, CT F628)
26 Middle English: Phonology And the same could even happen to full lexical nouns like men in (8): (8)
Yet ˈseen men ˈwel by ˈresoun, ˈdouteˈlees, (c.1390 Chaucer, “The Second Nun’s Prologue”, CT G16)
Thus, ME utterance rhythm was not simply an expression of lexical stress patterns. Instead, the relationship between word stress and rhythmic prominence was a matter of probabilities. Of course, lexically stressed syllables in polysyllabic words could count on winding up prominent in utterances, and their unstressed neighbors could normally count on showing up weak, but apart from that, rhythmic prominence was variable and context dependent, most strongly so in the case of monosyllabic prepositions, conjunctions and other grammatical word forms. That Middle English utterance rhythm preferred strong and weak syllables to alternate regularly had the effect that the intervals in which prominence peaks followed one another were rather regular also in terms of their actual duration. Of course, some feet would be shorter than others, but even when a stressed monosyllable was immediately followed by another stress, as crist is by cumeth in (9) it was often heavy and long. (9)
|sw |sw |sw |s |sww |sww | ˈure | ˈlouerd | ˈihesu | ˈcrist | ˈcumeð of | ˈheuene our lord Jesus Christ comes from heaven (c.1225 Trinity Homilies XII 67)
On the other hand, when feet contained more unstressed syllables than the norm, those unstressed syllables would often be light and short, so that the duration of a foot rarely deviated too far from the norm. Therefore, most Middle English utterance feet would last about equally long and English speakers would come to expect the time spans between rhythmic peaks to be fairly regular. It makes sense to assume that Middle English utterance rhythm was at least on its way to becoming as isochronous as Modern English rhythm is. We shall see that this assumption helps to understand some developments in segmental phonology.
4 Segmental phonology Although the developments which English sounds underwent during the roughly five hundred years of Middle English were naturally numerous and diverse, some generalizations can be made. They help to structure the account and suggest a certain causal coherence behind the individual events. (1) In unstressed syllables there was a clear tendency for segments and clusters to be weakened. (2) Segment duration lost much of its contrastive function, and the phonological distinctions it had encoded came to be expressed by quality contrasts. (3) Consonants were reduced more strongly than vowels.
4.1 Reductions in unstressed syllables Reductions in unstressed syllables are a natural phenomenon. In unstressed syllables phonetic contrasts are diminished, and phonologically distinctive features are
405
406
IV Middle English expressed imperfectly. Since Middle English words had stress lexically fixed on one of their root syllables, most other syllables would often be backgrounded, some of them practically always. Produced with comparably little articulatory effort, segmental contrasts would be perceived less well. In Middle English these processes had far reaching historical effects. It seems as if speakers tended to forget that the sounds they produced and perceived in regularly unstressed syllables were reduced variants of targets that had originally been meant to be more distinctive. These reduced sound variants then came to be memorized as primary articulatory targets themselves and ousted their stronger ancestors from the positions they had formerly occupied. The sounds which were most systematically affected by weak syllable reductions of that type were vowels (cf. Luick 1964 [1914–21]: 297–309, Lass 1992: 77–78.). Already in Old English long vowels had been shortened in weak syllables, and in Middle English also the quality contrasts among them were neutralized. Eventually the only distinction that remained stably established in most weak syllables was that between short /i/ and a schwa-like default vowel, and even in their case, spelling evidence ( ~ ‘king-GEN’, ~ ) and later developments suggest that in late Middle English they were also becoming exchangeable. The schwa-like default vowel itself had resulted from a merger of Old English /e/, /u/, /o/, and /a/. Evidence for that merger comes mostly from spelling, which for a period was characterized by much and rather unsystematic variation in the representation of the affected vowels: for example, , , , and later also and in the second syllable of heofon ‘heaven’ from OE /o/; and in infinitives like , ‘to bind’ from OE /a/; and for OE sunu, etc.). Eventually seems to have been adopted as the default letter for the default vowel. It is important to be aware that not all weak syllables were equally affected by vowel reductions, and that the relation between lack of stress and phonological reduction was not as mechanically causal as it might first look. Rather, some syllables – particularly inflectional endings – were affected more strongly than others. In them, phonological weakening went hand in hand with a loss of morpho-semantic function (see Wełna, Chapter 27). Also, in post-tonic root syllables, vowels were generally reduced strongly, particularly if they followed a lexically stressed syllable immediately (as in thusnd ‘thousand’, or heofn ‘heaven’). In prefixes, on the other hand, reduction does not seem to have gone that far, and did not always produce schwas from OE /a/, /e/, and /o/. For example, /o/ seems to have remained stable in the prefix for-, and in the prefix be- (from OE bī -) (before) the vowel may have been a schwa for some time, but appears to have been raised and fronted again to /i/. Also, in suffixes like -hood(e), or -ness(e), unreduced vowels (i.e. /oː/ and /e/) seem to have survived. Here, the relative stability of vocalic contrasts may have to do with the transparency of the involved morphs, their relative semantic weight, the fact that they could bear secondary stress in certain formations (e.g. ˈbiscopˌhood, ˈholiˌnesse), or a combination of all these factors. Perhaps even more consequential than the quality reduction of stronger vowels was the wholesale loss of unstressed syllables through schwa deletion, which affected large parts of the Middle English lexicon (see Minkova 1991). Schwa deletion started as an optional post-lexical process, which was at first restricted to word-final schwas in hiatus (i.e., when the word that followed them began with a vowel). Then it seems to have
26 Middle English: Phonology been gradually extended to word-final schwas in general, and finally to schwas in closed syllables, except where the deletion would have yielded a phonotactically ill-formed cluster. While in closed final syllables schwa deletion remained optional until the end of the Middle English period, it became obligatory in open syllables. In absolute word final position schwa is likely to have disappeared also from lexical phonological representations by the 15th century. The stages of this development can be charted roughly as shown in (11) – (13): Stage 1: ə → Ø / σs(σ) X__#V (11)
Himm ˈsholld(e) onn ˈeorþe ˈshæwenn (c.1180 Orrmulum H 876)
Stage 2: ə → Ø / σs(σ) X__# (12)
Vor ˈwan(e) þu ˈsittest ˈon þin(e) ˈrise (c.1200 Owl and Nightingale 894)
Stage 3: ə → Ø / σs(σ) X__Y, where XY is phonotactically well formed. (13)
For, ˈlording(e)s, ˈsith I ˈtwelf yeer ˈwas of ˈage, ˈThonk(e)d be ˈGod, that ˈis eˈtern(e) on ˈlyve, (c.1390 Chaucer, “Wife of Bath’s Prologue”, CT D4–5)
Of course, there were overlaps and much sociostylistic variability in what looks like a clean sequence in the representation above. Also, the development was complicated by other factors, including morphological ones: word-final schwas often represented the reduced residues of inflectional suffixes, which were in the process of losing their morpho-grammatical functionality. As they did, they seem to have fused with lexical stems and been reinterpreted as parts of base forms. In such cases, Xə and X forms seem to have become allomorphs. The Orrmulum provides many examples of such variation, e.g. (14): (14)
a. 7 brohht to grund 7 underfot (c.1180 Orrmulum H 11773) b. to brinngenn him to grunde (c.1180 Orrmulum H 12547)
Even word forms in which a final schwa was morphologically unwarranted could have one inserted, as in (15): (15)
forr he maʒʒ been swa grimme mann (c.1180 Orrmulum H 7174)
Being rhythmically motivated, schwa deletion reflected subtle constraints from utterance rhythm also in the way it was implemented. For example, it was delayed in contexts where it would have resulted in a clash of prominence peaks. Thus, disyllabic adjectives appear to have retained final schwas longer in attributive position, where they would often have been followed by nouns with initial stress (þæt ˈblake ˈsmoke vs. his ˈberd was ˈblak; cf. Minkova 1991: 178–184). That the choice between realizing or deleting schwa in final syllables was governed by rhythmic preferences becomes most obvious in poetic texts, of course, but poetry merely enhanced general tendencies in that respect.
407
408
IV Middle English On the consonantal side, the most noticeable reduction process affected nasals in the codas of unstressed syllables, where they were often deleted. These deletions were not completely systematic, however, and fully implemented only in nominal and adjectival case endings, and later in verbal infinitives ( findan > finde) and plural forms (wenten > went). In strong past participles (e.g. bundon > bounde, but: broken), they were sporadic, and even rarer among full lexical words (e.g. mæʒden > maide). Other consonant reductions made voiceless fricatives voiced in the codas and, to a lesser extent, in the onsets of weak syllables. This affected the final /s/ in nominal plurals and genitives, but also the codas of grammatical morphs that normally wound up unstressed, e.g. was, his, as, with, or of. In the prosodically stronger onset position, voiceless /s/ and /f/ remained stable even in unstressed syllables, however, while /θ/ was voiced to /ð/ in this, these, those, they, them, etc. Glottal /h/ was deleted in onsets like those of hit > it, or hem > ’em (the 3P PL pronoun).
4.2 Segments in strong positions While weak syllable reductions clearly reflect the effects of lexical word stress and rhythmically motivated backgrounding, unifying factors behind the developments that Middle English sounds underwent in other phonological environments are harder to detect. (For the segmental phonology of Middle English, see also Jordan 1934; Jones 1989: 94–190; Luick 1964 [1914–21]: 323–488; Lass 1992: 39–83; or Mosse´ 1952.) Nevertheless, the following overview will attempt to impose some order and narrative coherence on the events. First, however, a brief overview of the Early ME phoneme inventory is given in Table 26.1. Table 26.1: Early Middle English vowel phonemes Short monophthongs Front
Back
/i/ /y/ sitten ‘sit’, fyllen ‘fill’ /e/ /ø/ setten ‘set’, heofon ( lauerd ‘lord’), /hr/ (hrydcʒ > ridge), or /hn/ (hnut > nut), and seems to have been unstable even when it was the only consonant in a stressed onset, as occasional spellings like algen ‘hallow’, ate ‘hate’, or reversed herðe ‘earth’ suggest. In codas, [x]/[c¸] remained stable until the 15th century, when it came to be lost (plough, knight), or merged with /f/ (rough). Other consonants were also lost. Thus, /w/ disappeared in onset clusters like /sw/ (e.g. such < swylc, sword /sɔːrd/), /l/ sometimes in codas before other consonants (each < /æːltʃ/, such < swylc), and even sporadic precursors of /r/-loss seem to have occurred (e.g. ass < ars). Furthermore, occasional metatheses like bird < bryd, horse < hros, or through < þurh suggest that /r/ was sometimes felt to belong to the vocalic syllable nucleus. Although obstruents were more resistant to weakening, they were not completely immune either. Thus, word final /mb/ and /ng/ clusters were simplified to /m/ and /ŋ/ as in comb, climb, or long, and intervocalic /d/ sometimes was weakened to /ð/, as in father (< fader) or weather (< weder).
413
414
IV Middle English
5 Summary This survey has necessarily been simplified and has left many details unreported. Yet, the thrust of its argument is likely to remain valid if enriched with more data, and if socioregional variability is taken more strongly into account: many properties of the Middle English sound system and its development can be related to the preference for rhythmic isochrony and to the fact that Middle English words had lexically fixed stress positions. The combination of these factors might explain the decay of quantity distinctions among consonants and vowels, the weakening processes affecting unstressed syllables, and also, albeit less directly, the emergence of new diphthongs. It is even tempting to speculate that also the general tendency of consonants to be weakened more strongly than vowels may be relatable to the requirements of rhythm: consonants are generally less flexible than vowels as far their phonetic duration is concerned, and isochronous rhythm obviously depends on such flexibility. Tempting though it is, this interpretation is clearly in need of further substantiation.
6 References Dresher, B. Elan and Aditi Lahiri. 1991. The Germanic foot: Metrical coherence in Old English. Linguistic Inquiry 22: 251–286. Dziubalska-Kołaczyk, Katarzyna. 2002. Beats-and-Binding Phonology. Frankfurt: Lang. Jordan, Richard. 1934. Handbuch der Englischen Grammatik. Heidelberg: Winter. (Rev. edn. by Charles Matthes, 1968.) Jones, Charles. 1989. A History of English Phonology. London: Longman. Kurath, Hans. 1978. The loss of long consonants and the rise of voiced fricatives in Middle English. Language 32: 435–445. Lass, Roger. 1992. Phonology and morphology. In: Blake, Norman (ed.), The Cambridge History of the English Language. Vol. II. 1066–1476, 23–155. Cambridge: Cambridge University Press. Luick, Karl. 1964 [1914–21]. Historische Grammatik der englischen Sprache. Stuttgart: Tauchnitz. Minkova, Donka. 1982. The environment for Middle English Open Syllable Lengthening. Folia Linguistica Historica 3: 29–58. Minkova, Donka. 1991. The History of Final Vowels in English. Berlin/New York: Mouton de Gruyter. Minkova, Donka. 2006. Old and Middle English prosody. In: Ans van Kemenade and Bettelou Los (eds.), The Handbook of the History of English, 95–125. London: Blackwell. Mosse´, Ferdinand. 1952. A Handbook of Middle English. James A. Walker (trans.). Baltimore: Johns Hopkins University Press. Ritt, Nikolaus. 1994. Quantity Adjustment. Cambridge: Cambridge University Press. Stockwell, Robert P. 1986. Assessments of alternative explanations of the Middle English phenomenon of high vowel lowering when lengthened in the open syllable. In: Roger Eaton, Olga Fischer, Willem Koopman, and Frederike van der Leek (eds.), Papers from the fourth International Congress on English Historical Linguistics, 125–134. Amsterdam/Philadelphia: John Benjamins.
Nikolaus Ritt, Vienna (Austria)
27 Middle English: Morphology
415
27 Middle English: Morphology 1 2 3 4 5 6
Introduction Nominal system Verbal system Remarks on word formation Major problems in Middle English inflectional morphology References
Abstract The central part of the chapter contains a succinct account of ME morphology in the roughly four centuries following the Norman Conquest. Attention is focussed on inflectional morphology, especially on the gradual decay of inflectional paradigms which led to the disintegration of the system of grammatical markers. The most important morphological processes of the period include (1) the loss of the majority of oblique cases, generalization of the plural marker -s, and substitution of semantic for grammatical gender in nouns; (2) the loss of inflections and the initial stage of the spread of periphrastic comparison in adjectives; (3) the spread of the pronouns she, they, them, their; and (4) the loss of inflections, decay of the subjunctive mood, and substantial disintegration of the ablaut system in verbs (i.e. the shift from strong to weak). All these processes are viewed against the background of regional variation, with special reference to the dialects which contributed to the formation of the morphology of Standard English. Due attention is given to the impact of phonological factors on the morphological processes. The final sections contain a brief presentation of ME word-formation types, including influences from French in the post-Conquest period, and a critical review of more important contributions to the field of ME morphology, from the end of the 19th century to our times.
1 Introduction Similar to the majority of ME grammars (Brunner 1963; Fisiak 1996) or handbooks of morphological change (Bammesberger 1975) attention is here focussed on inflectional morphology, whose evolution, apart from the processes of analogy, levelling, and perhaps contacts with Scandinavian, was to a large extent determined by phonological changes in English between c.1100–1500. Here belong especially the reduction of vowels in unstressed syllables, which triggered the loss of grammatical gender and levelling of the article forms, case markers in nouns (except genitive), adjectives, and, partially, pronouns. The most spectacular changes affected the adjective, which lost the categories of gender and case, also abandoning the distinction between the strong and weak declension. Verbs proved more resistant because, unlike nouns and adjectives, their inflectional markers contained obstruent consonants, i.e. sounds not subject to vocalization, as in PRES SG 2P -st, 3P -eþ (-eth) or PAST -ed, while the nasal sonorants -m
Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 415–434
416
IV Middle English (> -n) and -n, frequently found in the nominal endings, vocalized and were ultimately dropped. With reference to time division, ME inflectional morphology shows two distinct types. While texts from before 1300 still retained conservative features, including inflections (e.g. Poema Morale, which still shows DAT SG -e), the post-1300 morphological system became disintegrated, with only verbs preserving inflections; cf. Brunner (1963: 1). In what follows the term “nominal” is used with reference to the declined parts of speech and to adverbs.
2 Nominal system The complex OE grammar of the nominals (see von Mengden, Chapter 18) became simplified following the reduction of case/number distinctions in nouns, adjectives, and determiners. This was either due to phonological processes (reductions in unstressed syllables) or to analogical transfer of the strong markers (GEN -es, PL -as) from the masculine to feminine and neuter declensions. Nouns and adjectives lost DAT and ACC, while GEN SG and PL markers suffered levelling. These processes were accompanied by a gradual replacement of grammatical by semantic gender.
2.1 Nouns Although the old classification of nouns based on the stem structure requires a revision (cf. Kastovsky 1988), for convenience in what follows, reference is made to the traditional classification. As said earlier, inflections in the period under investigation fall into two distinct types: (a) with inflectional markers considerably levelled but still retained (c.1100–1300) and (b) with such markers practically eliminated (1300–1500).
2.1.1 Number and case The system of the plural marking began to lose transparency even in Late Old English when endings tended towards levelling, so that PL -as of the masculine a-declension began to spread to minor declensions. The change seems to have originated in the North, where the weak ending -(e)n was dropped and the former strong endings -es (GEN SG) and -as (PL) began to be also attached to non-masculine nouns. Nominative/accusative singular. In the original a-/-o¯-declensions NOM/ACC SG was usually unmarked, but word-final -e and -u are sometimes found: -e retained in ende ‘end’ (OE ende) but added in fowle (OE fugol ) ‘fowl’, etc. Some former members of the wa- (MASC/NEUT) and wo¯-declensions (FEM), i.e. nouns ending with -u, modified this vowel to -e, as in bale (OE bealu) or created nominatives based on the obliquecase forms with medial -w-, e.g. clawe (OE cle¯a-SG.FEM) ‘claw’. Other members of the declension either preserve (sno¯w < OE sna¯w) or drop -w- (kne¯ < OE cne¯o, -wes ‘knee’). Likewise behave nouns having stems in -i (giste/gest < OE giest-MASC ‘guest’) and in -u (dure/do¯r < OE duru-FEM ‘door’). The nominatives in -a (MASC), -e (FEM/NEUT) (ACC -an MASC/FEM, -e NEUT) of the weak nouns also lost distinctions after they merged as -e [ə]. Nouns from minor consonantal declensions, like bro¯ther (r-stem), fie¯nd (nd-stem), lamb (iz/az-stem), mo¯nth
27 Middle English: Morphology ( þ-stem) retained their nominative forms and so did nouns of the root-consonant stem declensions, like foot, man, tooth, woman (MASC), book, cow, go¯s ‘goose’, lous ‘louse’, mouse (FEM), etc. (For their plural forms see below). Genitive and dative singular. Of these two markers, only the genitive survived in Middle English, while the dative with either strong -e or weak -an, merged into -e [ə], dropped soon. The obstruent in the genitive prevented the loss of -es (MASC/NEUT) which thus could spread in dialects. The early occurrence of an apostrophe is registered in Robert of Brunne (1338): kynge’s William broþer (cf. Graband 1965: 114), but such forms increased after Middle English. The synthetic forms with -es competed with analytic constructions containing the preposition of, frequently found with nouns denoting abstract entities or things (c.85% around 1300; cf. Fries 1938). As a consequence, the inflected genitive became confined to personal nouns, including proper names. Nominative and accusative plural. While the South favored weak plurals in -en, the originally MASC PL -(e)s (OE -as) spread to nouns of the other genders, first in the North and North Midlands, affecting on a larger scale nouns in the South only in the 13th (Southwest) and 14th (Southeast) centuries. Also loanwords from French regularly received -(e)s. These two types, with -es and with -en (South, The Ancrene Riwle c.1225, William of Shoreham c.1315, Ayenbite of Inwyt c.1340), are exemplified below: adduced are only forms with -es or -en extended to words with an originally different plural marker or unmarked): 1. -es (a-NEUT) ge¯res ‘years’, hu¯ses ‘houses’, sweordes ‘swords’; (n-FEM) lefdies ‘ladies’; (r-MASC) uederes ‘fathers’; (a-NEUT) ga¯tes, heauedes ‘heads’, wordes ‘words’; (o¯-FEM) halues ‘halves’; (nd-MASC) vrendes ‘friends’. 2. -en (o¯-FEM) soulen ‘souls’, zennen ‘sins’, wunden ‘wounds’; (a-NEUT) dye¯vlen ‘devils’; (u-MASC) sunen ‘sons’; (r-MASC) breþren ‘brothers’; (r-FEM) dogtren ‘daughters’; (-iz/-az-NEUT) children, eiren ‘eggs’, etc. Although typical of the South/South Midlands, the diffusion of the weak plural lost impetus because of the ongoing vocalization of -n. Originally in common use, weak -en frequently coincided with strong -es, e.g. a¯pen/a¯pes, arwen/arwes ‘arrows’, etc. Still, the rivalry continued with a diminishing intensity into the 15th century, when only remnants of the weak declension in -en survived. The root-consonant stems (mutated) plurals successfully survived the regularizing processes, continuing to express plurality through vowel change ( fe¯te ‘feet’, geese, lı¯ce, men, mı¯ce, teeth, women), but several items assumed -es endings, cf. bookes (OE be¯c) ‘books’, borughes (Trevisa; OE byrh) ‘boroughs’, gootes/gaytes (OE gæ¯t) ‘goats’, etc. The plural bre¯c (OE bro¯c SG) developed a singular meaning, while several mutated plurals failed to survive beyond Old English (æ¯c ‘oaks’, *hnyte ‘nuts’, tyrf ‘turfs’). OE
417
418
IV Middle English cy¯ ‘cows’ developed a regular plural only after ME, while irregular kı¯ne ‘cows’ still survives in dialects. Although former strong neuters continued to use unmarked plurals (cf. chı¯ld ‘children’, hond ‘hands’, hors ‘horses’, winter ‘winters’), they gradually adjusted to the dominating pattern with -s as the plural marker. Today, only deer, sheep, and swine take no -s in the plural. Cf. also Roedler (1911, 1916) and Newman (2008). Genitive and dative plural. The reduction of vowels and nasals in endings eliminated the old GEN PL (-en)-e < OE (-en)-a, which survived only temporarily in the South. Very similar was the fate of the DAT PL -en (OE -um), which vocalized and disappeared. As a consequence of these processes the genitive and dative plural ceased to be distinguished.
2.1.2 Gender The loss of grammatical gender, a consequence of the decay of inflections, was a functional change which began in the North around 1200, reaching the South some two centuries later. That development coincided with the rise of semantic, or natural gender, whose first signs are evident in Late Old English. According to Moore (1921), the two events involved two separate developments. The loss of grammatical gender is reflected in the Peterborough Chronicle, whose three segments show how genderdistinctive inflections were eliminated. The explanations of the process vary. Thus, Clark (1957) attributed the loss to a masculinizing tendency, while Markus (1988), to the rise of the uninflected definite article (see Jones 1988).
2.2 Adjectives The immensely complex inflections of the OE adjective simplified in Middle English when it lost practically all inflectional markers, eliminating the distinction between the strong and weak declensions and losing the grammatical categories of gender, number, and case. The only surviving category was that of degree.
2.2.1 Inflection Adjectives lost endings due to either reduction of unstressed syllables (-en, -ene-, later -e) or to analogical influences caused by the paradigm levelling in the South/Midlands (Fisiak 1996: 76–79). Although several oblique case markers continued to be used in the South, the only ending which survived, -e, was for some time attached in the nominative of the former strong and weak declensions, e.g. black(-e) ‘black’, smal(-e) ‘small’, etc. The loss of inflectional -n in the early 13th century resulted in the merger of -e(n) with [ə], which was subsequently lost, first in the North. The above concerned non-poetic language: in poetry, [ə] in adjectives survived into Late Middle English, extensively employed for reasons of metrics by Chaucer and other poets. Minkova (1990: 330) writes that “final -e as a singular adjectival grammatical marker in Late Middle English appeared intact only as a result of the complex interaction between syntax, morphology, and the requirements of the speech rhythm”. Loanwords, especially from French, are not distinguished morphologically from the native adjectives. Case inflections on the comparative and superlative forms were also lost.
27 Middle English: Morphology
2.2.2 Comparison The comparative suffix -re and superlative -est, the continuations of OE -ra/-ost, survived with only minor phonological modifications; cf. ME bold, bolder, boldest. An essential change was the replacement of the irregular mutated forms of the comparative and superlative by the stem of the positive, e.g. short, short-er, short-est (OE sceort, scyrt-ra, scyrt-est), etc. Such levelling affected adjectives with the original mutated comparative and superlative stems like bro¯d ‘broad’, o¯ld, yung ‘young’, etc. (OE bræ¯d-ra, -est, ield-ra, -est, gieng-ra, -st, etc). The suppletive comparison continued in spite of a tendency to level roots in the three forms (e.g. badder, baddest, of badd(e), coexisting with wurs(e), wurst). The 13th century saw the first signs of periphrastic comparison, when mo¯(re), mo¯st began to modify adjectives to form analytic degrees. This type gained a stronger position in the two following centuries; cf. Gonza´lez-Dı´az (2008) and Janecka (2008).
2.3 Numerals Although phonological change affected the forms of ME numerals their system remained practically unchanged, with only minor modifications affecting cardinal and ordinal numbers. Like adjectives, also the cardinal numerals lost inflections early. The numeral ‘one’ developed two variants, stressed o¯n ‘one’ (sometimes o¯ and Southern oo before consonants), with the semivowel [w] attached initially in the 15th century (cf. Southwestern wo¯ne), and unstressed an. Ultimately, the unstressed variant lost -n before vowels, which produced the present-day indefinite article with preconsonantal a and prevocalic an. The original threefold gender distinction of the numeral denoting ‘two’ (OE tweȝenMASC, twa ¯ -FEM/NEUT, tu¯-NEUT) was abolished, although the former masculine form twayn continued into the 15th century without gender distinction. But it was the original feminine/neuter two¯ (OE twa¯) which became a standard form. The remaining numerals survived as thre¯ ‘three’, foor ‘four’, five, six, seven, eyght(e), nı¯ne, te¯n, elleven, twelve. The suffix -te¯n was attached to the numerals 13–19 (þrette¯ne, fourte¯ne, fifte¯ne, sixte¯ne, sevente¯ne, eiȝtete¯ne, niȝente¯ne), while the numerals 20–90 attached the suffix -ti (twenti, þretti, fourti, fifti, sixti, seventi, eiȝteti, niȝenti). The highest numbers were hundred, thousand, and millioun ‘million’ (< Fr., 14th century). The dialectal distinction between the ordinal forms of ‘first’ continued in Middle English: first (East Midlands/Northern), furst (West Midlands/Southwestern), uerst (Kentish). The most spectacular innovation, the replacement of other by Fr. secounde ‘second’, occurred in the 13th century. The metathetic form therd(e) replaced thridde a century earlier. During the period, higher ordinal numbers except fifte ‘fifth’, sixte ‘sixth’, ellefte ‘eleventh’, twelfte ‘twelfth’ developed the regular marker -þ, later -th (fourþe, sevenþe, eiȝteþe, niȝeþe, te¯þe). The ordinal numbers thirteenth to nineteenth had the suffix -te(o)þe, -tiȝeþe later replaced by -tenþe ‘-teenth’, while the numerals ‘twentieth’ to ‘hundredth’ attached the suffix -ti(ȝ)(e)þe . Ordinal numbers higher than ‘100’ were virtually absent in Middle English.
2.4 Adverbs To form adverbs the suffix -e was attached to adjectives. Because of the reduction of unaccented vowels in the word-final position such adverbs were distinguished from
419
420
IV Middle English the respective adjectives only through their place in a phrase; cf. la¯te, swe¯te ‘sweet’ (ADJ/ etc. Consequently the suffix -liche (OE -lice) transformed as -ly (Northern -lik) became the principal adverbial marker, cf. fully, soothly, wı¯dely, etc. Levelling affected adverbial comparative/superlative degree forms, e.g. strong-er : strong-est, the suffixes -er, -est replacing OE -ra, -ost/-est (cf. streng-ra : streng-est). However, forms without a suffix, like leng ‘longer’, mo ‘more’, bet ‘better’ continued to be used.
ADV),
2.5 Pronouns The earlier system embracing personal, possessive, demonstrative, interrogative, and indefinite pronouns acquired a new type: reflexives. Simultaneously, the interrogatives who, what, etc. began to function as relative pronouns. The pronoun was the only nominal part of speech which retained a fully-realized category of case.
2.5.1 Personal The ME system of pronouns became modernized due to external influences (borrowing) but also as a consequence of language internal processes. Thus, ich ‘I’, frequently dropped in the Midlands (12th century) especially before a following consonant. The surviving became capitalized in the 14th century, and the vowel became lengthened before the Great Vowel Shift. Chaucer’s verse contains both types, ich being less frequent, while I only is found in his prose. In the plural, vowel shortening affected the object case u¯s (Ormulum, c.1200), but the spelling in the Paston Letters (15th century, East Anglia) testifies to the survival of long [uː] beyond Middle English. Standard we¯ substituted for the dual wit ‘we two’ in the 13th century. The transformations of the 2P pronouns were determined by sociolinguistic factors: a tendency to employ one pronominal form for the singular and plural. Originally, thou was used when referring to inferiors, and ye¯ (also singular from the 13th century), when addressing superiors. The current form you, perhaps the contamination of ye¯ and ou (object), is first recorded in the 14th century used then as the plural object case, the nominative still being ye¯. Curiously, the nominative singular you is found in the 14th century, i.e. earlier than the respective plural. In the 13th century the dual yit ‘two of you’ was replaced by ye¯. While the oblique feminine (her, object case) and neuter (h/it) forms are only slightly modified, the merger of masculine he¯ and feminine he¯o ‘she’ (e¯o > e¯) abolished the functional contrast between the two pronouns. This triggered the rise of feminine sche¯ ‘she’ in the 13th century, perhaps influenced by Northern scho¯ (Ormulum ȝho). The conservative character of the West and South is reflected in the survival there of the old h-forms into the 15th century. Analogous functional reasons seem to have underlain the replacement of the plural forms hi(e/hio/he/ho (oblique hem/hom/him) by the Scandinavian loanwords ðeȝȝ /thei ‘they’ (oblique thaim/them) because h-forms were phonologically similar to the masculine and feminine pronouns. The new th-forms appeared in the North in the 12th century, spreading a century later over the Midlands and South, where both types are found in the 14th century. In London, forms with initial h-, like hem ‘them’, still survived a century later. The standard distribution in Middle English was as follows: th-forms in
27 Middle English: Morphology the North, they vs. her, hem in the Midlands (Chaucer), h-forms only in the South. The spread of the oblique th-forms in the South was slower and they were for a long time confined to documentary and late texts (cf. Stenroos 2005).
2.5.2 Possessive The 1P possessive developed two forms, stressed mı¯ne and unstressed mı¯ ‘my’ (now an adjective). Originally their distribution was governed by syntactic factors, mı¯n appearing before vowels, mı¯ before consonants. In the 14th century the plural forms u¯r/ou¯r (now an adjective) attached –s (ou¯res, now ours, a pronoun). The form was in all probability influenced by the 3P SG possessive his. Like other duals, uncer ‘of the two of us’ was replaced by the standard form our. Like the 1P SG, the 2P possessive also distinguished forms with and without -n (thı¯ne, thı¯ ‘thy’), earlier in accordance with stress conditions, and later depending on the position before vowels or consonants. The first use of your in the singular occurs at the end of the 13th century, while the pronoun yours was coined a century later. The dual incer ‘of the two of us’ ceased to be used in later Middle English. The possessive his continued as a form common to the masculine and neuter, while hir/her developed the pronominal possessive form hers (15th century) The new thpronouns are recorded in the 13th century, e.g. þeȝȝre (Ormulum). The distribution of possessive h- and th-forms in the plural was analogous to those of personal hi/they (see Section 2.5.1). The rare instance of the possessive ðayres ‘theirs’ is found in the Cursor Mundi (North).
2.5.3 Demonstrative The complex five-case inflections of the OE demonstrative became simplified, especially in the North and East Midland, while the South and West continued to use a four-case system until the 14th century. The replacement of the initial s- (OE se¯, se¯o) by þ-/th- was an essential step towards establishing the definite article the, a form recorded in Northumbrian Old English. Wright (1928: 167) observes that “In the northern and East Midland dialects the uninflected nom. masc. and fem. form þe had come to be simply the by about 1150, and almost everywhere else by about 1300.” But before 1300, in the majority of English dialects the former neuter demonstrative þat ‘that’ prevailed before vowels, instead of the. The true demonstrative pronouns are this, that (former NOM/ACC NEUT) and the¯se (former MASC/FEM NOM SG), tho¯se (PL). In general, such pronouns were first adopted in the North and East Midland, later spreading to the South. The “remote” plural those appears in Wyclif (Midlands) long before the general acceptance of the form in the 15th century.
2.5.4 Reflexive Although in Old and in Early Middle English the pronoun self was used for emphasis after personal pronouns, in later Middle English it began to stand after 1/2P SG possessives, gradually merging with them into compounds, cf. myself, thyself, yourselfe ‘yourself’. Similar combinations with 3P SG masculine possessive hisself were replaced by himself. The form it self (OE hit self ) is first recorded in the 12th century.
421
422
IV Middle English The respective plural forms of the reflexive originated very late, the last one being the 2P PL *yourselve, recorded only after Middle English, while the 1P PL ourselfs/our selves ‘ourselves’ and themself ‘themselves’ emerged in the 15th century.
2.5.5 Interrogative, relative, and indefinite The reduction of cases also affected the interrogative pronouns who (MASC/FEM) and what (NEUT). As a consequence the emerging inflectional paradigm did not differ much from that functioning in current English (who, whom, whose, what). The old instrumental case hwı¯ (OE hwy), reinterpreted as an interrogative adverb, joined words like when, where etc. In addition, two more interrogatives were current: hwilc/ which OE hwylc) and hweþer/whether (OE hweþer), both exhibiting strong dialectal variation. A completely new system of relative pronouns emerged after the historical relative þe went out of use in the 13th century. Its function was taken over by the interrogative pronouns who¯se, who¯m and which, while who¯ used as the relative joined other pronouns only in the 15th century. Middle English also exploited þatt ‘that’ as a relative pronoun (e.g. the Ormulum). Several adverbial relatives like whe¯r ‘where’, whe¯reas, when were also in use. ME indefinite pronouns were chiefly reflections of their OE counterparts. They can be represented by items like all, any, both, each, either, elles ‘else’, eny body ‘anybody’, euery ‘every’, many, neither, no¯ne, o¯ne, o¯ther, some, such, etc. As said earlier (see Section 2.3.1), the pronoun/numeral an (‘one’) began to serve as the indefinite article.
3 Verbal system The grammatical categories of OE verbs, i.e. person, number, tense, mood (see von Mengden, Chapter 18), were retained in Middle English. Also, the division of verbs into weak and strong remained valid, although their distribution changed because of many strong verbs attaching the suffix -ed to mark the past tense. The end of the period coincided with the incipient stage of the rise of the passive voice and two types of aspect.
3.1 Inflectional markers The inflectional paradigm of Middle English verbs was as follows: 1. present indicative: SG 1P -e, 2P -st, 3P -th, PL -en (all with -es in the North). Due to i-mutation strong verbs continued to exhibit fronted root vowels in the 2/3P SG. Subjunctive: SG -e, PL -en. Imperative: 2P SG Ø /-e, 2P PL -eth. 2. preterit indicative: sg. 1P -e (WEAK)/-Ø (STRONG), 2P -est (WEAK)/-e (STRONG), 3P -e (WEAK)/-Ø (STRONG), PL -e(n). Subjunctive: SG -e (North -Ø), PL-(e)n (all with the preterit marker -(e)d- attached after the root). 3. other markers: INF -e(n), PRESPRTC -end/-ind/-and (North)/-ing (Midlands), PASTPRTC - ed/-en.
27 Middle English: Morphology
423
The decay of inflections was initiated in the Northern dialect which lost the endings of the infinitive, 1P indicative, subjunctive, and imperative. The currently surviving 3P SG present ending -(e)s spread from the North. That dialect also contributed to paradigm levelling when the 2/3P SG present tense root vowels began to be replaced by the nonmutated vowels of the 1P SG. Likewise, the North removed the prefix ȝe- marking the past participle (which still survived as Southern i-/y-) but retained the suffix -en, serving the same purpose. Other dialects also contributed to the new inflectional paradigm. Verbs in the Midlands replaced the PRES PL -eth by -en, which spread to the South. The West Midlands invented the new PRESPRTC form -ing(e), perhaps a modified form of conservative -inde. The former may have been reinforced by the gerund in -inge.
3.2 Weak verbs Weak verbs formed their past/past participle forms with the dental suffix -d/-t. The predental -i-, lost in the North and East Midlands, remained in the South and West. The class of weak verbs grew following the adoption by many strong verbs of the past tense dental marker. Here, the traditional division into three classes is replaced by a classification into two groups, (1) verbs with the past/participle ending -ed (Northern -id/-yd), from OE -od(e) (original class 2 verbs) and (2) verbs with -d/-t- (the remaining verbs).
3.2.1 Class 1 This class constitutes the basis of the present-day class of regular verbs with -ed in the past tense (see Table 27.1). Here belong native and foreign verbs. Table 27.1: Class 1 weak verbs INFINITIVE acsien/ask call(en) chacen/chase chew(e) dı¯e(n) live love ne¯de/need pleie/play pro¯ve use(n) wend(e)
PAST/PAST PARTICIPLE (< (< (< (< (
t in the past/past participle INFINITIVE
PAST/PAST PARTICIPLE
bende ‘bent’ blend buylde ‘build’ lend rende ‘rent’ send spend
bent blent bild (PAST)/bilt (PASTPRTC) lent rent sent spent
Table 27.4: Class 2 weak verbs with root-final postvocalic -d/-t- with root-vowel variation INFINITIVE
PAST/PAST PARTICIPLE
ble¯de feede ‘feed’ hı¯de le˛e˛de ‘lead’ meet spe¯d ‘speed’ spre˛e˛de
bled fed hid led mett sped spredd (PAST)
27 Middle English: Morphology
425
Table 27.5: Class 2 weak verbs with root-final -t showing no variation of vowel length, now invariable verbs INFINITIVE cast cost cutt hit hurt put set shutte ‘shut’ slit spit thruste ‘thrust’
PAST/PAST PARTICIPLE (< ON) (< OFr.) (< ON) (< OFr.)
cast costed (PAST)/coste (PASTPRTC) cut hit hurt put set shut slit spit (later also with -a- in the PAST) thrust
Table 27.6: Class 2 weak verbs with suppletive bases (with root-vowel variation) INFINITIVE
PAST/PAST PARTICIPLE
bese¯che ‘beseech’ bring buye ‘buy’ catche ‘catch’ seek te˛e˛chen ‘teach’ think
besought (PAST) brought bought (PAST) caught sought (-k from inflected forms) taught thought
(< OFr.)
3.2.3 Weak verbs which developed strong forms Middle English demonstrated an evident trend towards reducing the number of strong verbs, which helped regularize inflectional paradigms (see Section 2.3). But simultaneously several weak verbs began to mark tense distinctions by differentiating root vowels and sometimes by attaching the past participle ending -en to the originally weak form. The trigger of such changes was analogous to that determining forms of the common strong verbs. Table 27.7 includes items whose new strong forms survived into Modern English, but excludes those which exhibited short-lived strong forms. Loanwords expressed tense difference through root vowel variation extremely rarely, cf. strive (< OFr.; PAST stro¯ve/strı¯ved). Table 27.7: Weak verbs which developed strong forms INFINITIVE
PAST/PAST PARTICIPLE
chide ring sawe sew shew/show stycke ‘stick’ we˛e˛r ‘wear’
chidde/cho¯de (PAST)/chidden (PASTPRTC) ringde/rongen (PL) (PAST)/runge (PASTPRTC) sawed/sew (PAST)/sawid/sown (PASTPRTC) sowed (PAST)/sewed/sowen (PASTPRTC) showed (PAST)/showed/showen (PASTPRTC) stiked/stacke (PAST)/sticked/sto¯ken (PASTPRTC) we¯red/wo¯re (PAST)/we¯red/worn (PASTPRTC)
(< ON ‘saw’)
426
IV Middle English
3.3 Strong verbs Strong verbs expressed the present/past tense distinction by using different root vowels. Continuing the OE pattern also the singular and plural preterit roots exhibited vowel differences, but a drift towards levelling the forms of the preterit/past participle led to their uniformity, with the vowel either from the preterit or from the past participle generalized. Ninety-seven OE strong verbs were lost in Middle English, while many others underwent levelling, first in the North, then in the South (Krygier 1994: 251). Also alternating consonants were subject to levelling in verbs, cf. fre¯se, fro¯se, fro¯zen (< froren), etc. The most important process which affected strong verbs was their shift to the weak class. Although only some ME strong verbs permanently established new fully weak forms, almost all of them attached dental preterit markers. Frequently confused were verbs in Classes 4 and 5 or those in Classes 6 and 7. The process of the strong-to-weak shift gathered speed in the 13th–14th centuries. The division of verbs into seven classes seems to hold, with each class exhibiting four subtypes (Wełna 1996: 120–121): 1. 2. 3. 4.
strong (STRONG; no weak forms); strong/weak (STRONG/WEAK; surviving as strong with occasional weak forms); mixed (MIX; both forms survive in Modern English); and the most advanced, weak/strong (WEAK/STRONG; tending to retain weak forms only).
The statistical data below concerning the relevant forms in Old English and those surviving now come from Brunner (B, 1962: 209–252,) and Krygier (K, 1994: 255–267)
3.3.1 Class 1 INF
-ı¯-
PAST
-o¯-/-i-
PASTPRTC
-i-
Strong Class 1, with its very transparent vowel alternation pattern, proved relatively resistant to the levelling trend and, as Hogg (1992: 152) writes, “it forms the basis of a fairly stable group of irregular verbs in Present-Day English”. Of the total number of strong 1 verbs in Old English (65 in B, 71 in K), 10 retain vowel alternation, while 13 exhibit mixed forms in their subsequent evolution. The most characteristic forms include rı¯de (STRONG); (a)rı¯se, bı¯te, drı¯ve, shı¯ne, slı¯de, smı¯te, strı¯ke, wrı¯te (STRONG/ WEAK). Mixed forms possess verbs such as rı¯ve, shrı¯ve, strı¯ve (< OFr.) (MIX), while other common verbs, e.g. glı¯de, wrı¯the (WEAK/STRONG), etc., abandoned vowel alternation as an indicator of the past tense. Because of their modified vocalism a group of OE contracted verbs like wre¯on ‘cover’, te¯on ‘accuse’, tended to join strong Class 2.
3.3.2 Class 2 INF
-e¯-~-u¯-
PAST
-e¯-~-u-
PASTPRTC
-o-
Having less transparent forms than strong 1 verbs, strong 2 exhibited two vowels (-e¯o-/-u¯-) in the OE infinitive, this variation being retained in Middle English. Following
27 Middle English: Morphology
427
the change -e¯oȝ- > -e¯ȝ- > -ı¯- certain verbs appeared with long -ı¯- in the infinitive (e.g. flı¯e ‘fly’). Of the total number of strong Class 2 verbs in Old English (51 in B, 60 in K), only 10 retain vowel alternation in Modern English, while 13 exhibit mixed forms at different stages of their evolution. The most important forms are che¯se (later cho¯se ‘choose’), fle¯o/flı¯/flyˆ, forbe¯de/forbid, fre¯se ‘freeze’ (with the levelling fro¯r-/> fro¯s- in PAST/ PASTPRTC) (STRONG/WEAK); cle¯ve ‘cleave (split)’ (confused with OE clı¯fan, hence a variety of strong and weak forms/(MIX)). The largest group is that of verbs eliminating strong forms, such as creepe ‘creep’, flee, le¯ȝe/lı¯e, sheete/sho¯te ‘shoot’ (original infinitives in -e¯), and bow, sho¯ve, sucke, sup (infinitives in -u¯-).
3.3.3 Class 3 (a)
INF
-i/ı¯-
PAST
-a~o¯-/-u-
PASTPRTC
-u-~-u¯-
(b)
INF
-e-
PAST
-a-/-u-
PASTPRTC
-o-
Strong Class 3 exhibits two distinct types, depending on whether their infinitives have roots with either prenasal -i- (-ı¯- before homorganic clusters) or non-prenasal -e-. The former subgroup includes verbs with a well-contrasted vowel pattern. Of the total number of strong 3 verbs in Old English (41 in B, 50 in K), as many as 22 retain root vowel alternation in Modern English with only 6 transferred to the weak class. Here belong strong verbs like sing, sling (< ON?), win; bı¯nd, fı¯nd, wı¯nd (the last three with a-preterits in the North) (STRONG) as well as begin, cling, drink, rinne (later run), schrinke ‘shrink’, sink, spinne ‘spin’, spring, sting, stink, swime ‘swim’, swing, wring, grı¯nd (STRONG/WEAK), while brinne/birne/burne ‘burn’, clı¯m ‘climb’ (WEAK/STRONG) develop weak forms surviving in Modern English. The group with the underlying vowel -e- contains verbs (47 in K) which dropped strong forms in Middle English, like berst/brust ‘burst’ (later an invariable verb), fight (STRONG/ WEAK), as well as swell (with PASTPRTC swollen; MIX), Other verbs exhibit only occasional strong forms, cf. bark, braid, carve, delve, help, melt, starve, swallow, swerve, thresh, warp, yell, yelp, yield (WEAK/STRONG). Practically all verbs in this subclass shifted to weak in the 12th–13th centuries, with only very few strong forms surviving longer.
3.3.4 Class 4 (a)
INF
-e¯-
PAST
-a-/-e¯-
PASTPRTC
-o¯-
(b)
INF
-u-
PAST
-o¯-~-a-
PASTPRTC
-u-
Verbs in this relatively small class (14 in B, 18 in K) are either lost or retain strong forms, although traces of the shift to weak are well documented. Their development reveals confusion with strong 5. In subgroup (a) exclusively strong is ste¯le ‘steal’ (STRONG), while other verbs occasionally use the dental suffix. Here belong verbs like beare ‘bear’, bre¯ke ‘break’, te¯re ‘tear’ (STRONG/WEAK), and shear (MIX). To subgroup (b) belongs come and its derivative become (STRONG), both exhibiting strong forms exclusively. Attention should be called to the emergence of the new preterit
428
IV Middle English cam of come, which originated in the Northumbrian dialect, later spreading to the Midland and South. The high frequency verb nim (OE niman) failed to survive beyond Middle English, replaced by Scandinavian ta¯ke with the same meaning (cf. Wełna 2005).
3.3.5 Class 5 (a)
INF
-e¯-
PAST
-a-/-e¯-~-e¯- (-o¯-~-o-)
PASTPRTC
-e¯-
(b)
INF
-i-
PAST
-a-
PASTPRTC
-e-~-o-~-i-
From the earliest time Class 5 verbs (26 in B, 30 in K) had a tendency to shift to strong 4. Thus, brecan ‘break’, which joined strong 4 in Old English, was now followed by spe¯ke ‘speak’, tre¯de ‘tread’, we¯ve ‘weave’, wre¯ke ‘wreak’. The reason for the shift could have been the resemblance of their present and preterit root vowels to those in strong 4. The typical representatives of the class were: (1) que¯th ‘-quethe’ (PAST quo¯th), e¯te ‘eat’ (STRONG); ge¯te/get, ge¯ve/ye¯ve/give, speake ‘speak’ (PASTPRTC spo¯ken for spe¯ken), tre¯de ‘tread’ (PASTPRTC tro¯den for tre¯den), we¯ve ‘weave’, se¯/see (STRONG/WEAK); beque¯the ‘bequeath’, kne¯de ‘knead’, fre¯t ‘fret’, me¯te ‘mete’, wey ‘weigh’, wre¯ke ‘wreak’ (WEAK/STRONG); and (2) bid (STRONG); lie, sit (forms often confused with those of weak sette ‘set’) (STRONG/ WEAK).
3.3.6 Class 6 (a)
INF
-a-~-a¯-
PAST
-o¯-/-o¯-
PASTPRTC
-a¯-~-a-~-o¯-
(b)
INF
-e-~-e¯-
PAST
-o¯-~-a-~-o¯-
PASTPRTC
-a(i)-~-ei-~-o¯-
Class 6 (31 in K) was characterized by a double set of root vowels, with much variation, so that Brunner (1962) distinguishes here four subgroups, of which one contains verbs contracted in Old English. This class showed a relatively strong tendency to shift to weak. To strong 6 belonged: (1) stand (also understand), wa¯de (STRONG); draȝen/draw, forsa¯ke, sha¯ke, swe¯re ‘swear’, ta¯ke (< ON; with numerous compounds, like underrta¯ken ‘undertake’, etc.) (STRONG/WEAK); gra¯ve (later displaced by engra¯ve, from OFr.), la¯de, sha¯ve, wa¯ke (MIX); a¯ke ‘ache’, ba¯ke, fa¯re, gnaw, laghe/laugh, sha¯pe (replacing sheppen), wash (WEAK/STRONG); (2) sle¯/slo¯/slay (STRONG/WEAK); hebbe/he¯ve ‘heave’ (MIX); fle¯n/flay, stappe/step (WEAK/ STRONG). Practically all these verbs developed forms with root variation, especially in dialects.
3.3.7 Class 7 (a)
INF
-e¯-
PAST
-e-
PASTPRTC
-e¯-
(b)
INF
-o¯-
PAST
-e¯-~-e-
PASTPRTC
-o¯-
27 Middle English: Morphology (c)
INF
-ow-
PAST
-ew-
(d)
INF
-e¯-
PAST -e¯-
(e)
INF
-a-
PAST
-e¯-(-e-)
429
PASTPRTC
-ow-
PASTPRTC
-e¯-
PASTPRTC
-a-
The specific character of strong 7 is determined by its containing former reduplicating verbs. As opposed to the simple division of these verbs into two classes in Old English, ME forms are split into at least 5 subclasses (66 in K, 7 subclasses), their vast majority shifting to the weak class. The strong past participle marker -en survives in verbs with the root sequence containing a vowel + . The particular subclasses include: (1) beate ‘beat’, le¯te/let (vowel shortened in the 13th century, but a long vowel in parallel use) (STRONG/WEAK); hew (MIX); dre¯d/dredde ‘dread’, leape ‘leap’, re¯d ‘read’ (chiefly weak in the 15th century), sche¯de/shed, sleepe ‘sleep’ (WEAK/STRONG); (2) ho¯ld (STRONG/WEAK); fo¯ld, swo¯pe/sweepe ‘sweep’, wo¯lde/weeld ‘wield’ (WEAK/ STRONG); (3) mowe ‘mow’, throw (STRONG); blow (‘move’), blow (‘bloom’) (both frequently confused, merging in the 15th century), grow, know (STRONG/WEAK); crow, sow (MIX); flow, row (WEAK/STRONG); (4) we¯pe ‘weep’ (STRONG/WEAK); (5) fall (STRONG/WEAK); walk, wax (‘grow’; WEAK/STRONG); hang (MIX). The fates of hang are particularly complex as apart from PAST heng the verb also developed forms with -a- and -u-. The present-day double forms of the past tense (hung vs. hanged) reflect the competition of the strong verb with weak intransitive hangian. The contemporary semantic distinction of the past tense forms was not yet established at the close of Middle English.
3.4 Preterit-present verbs This group of verbs retained the PIE perfect stems with the present meaning and consequently formed new preterit forms in Old English. In their evolution, preteritpresent verbs exhibited a drift towards modality (they survive as modals in Modern English). Their root vowels match the vocalism of the respective strong verbs (Classes 1, 3–6). Of the original twelve verbs, only seven survive in Middle English. Of the two verbs whose vocalism corresponded to that of strong 1 only owe (PAST ought/owed, PASTPRTC owne/owed) retained its strong position, while wit (PAST wiste, PASTPRTC witen/wist) hardly survived till the end of Middle English. In the 14th century owe became weak in the South, while ought acquired its present tense meaning a century earlier. Of the original four members whose vocalism corresponded to that of STRONG 3 only two retained their status: cunne/cann (PAST couthe/ coud/kowlde ‘can, could’) and durre/dear/dar/da¯re (PAST durst PASTPRTC dorren). The form cann reflects the 1/ 3P SG present, while the substitution of -d for -th in the past tense was analogical (from weak verbs). The -l- in could may be due to the analogy of wolde (> would),
430
IV Middle English the past tense of will. Also dare joined the class of weak verbs, while the two other verbs in this subclass, þurfan ‘need’ and unnan ‘grant’, ceased to be used very early. The only verb in strong 5, mowen/maiȝ/mai/may (PAST mihte/might, PASTPRTC mowe/ mowed) ultimately yielded the modal may (PAST might), reflecting OE mæȝ (1/3P SG PRES). The verb mo¯te (with the vocalism of strong 6) which developed the preterit mo¯ste, falling together with 2P SG mo¯st. It acquired the present tense meaning, and now the modal must survives without a past tense form.
3.5 Anomalous verbs All Old English anomalous verbs (be¯on/wesan ‘be’, do¯n ‘do’, ga¯n ‘go’, willan ‘will’) have survived into Middle English. The most important of them, be¯, continued to use suppletive forms in its inflectional paradigm: PRES SG 1P am, 2P art, 3P is, PL sind, are (PAST SG was, PL we¯re, PASTPRTC ibe¯/been). The are forms probably developed in the North or North East, spreading to the South and London in the 13th–14th centuries. The forms of the verb do¯ survive with only small modifications as do¯ (PAST did, PASTPRTC (i)do ¯ n). As with other verbs, the 3P PRES SG form showed dialectal variation: -(e)s in the North, -(e)th in the South. The most important innovation in the inflectional paradigm of go¯ was the curious replacement in later Middle English of the suppletive preterit e¯ode/yedd/yo¯de by another suppletive form, the preterit wente, of wende ‘turn’, while the past participle shows a regular development (go¯ne) (Wełna 2001). The verb wille/will survived unchanged into Middle English, with the preterit forms wo¯lde wulde would, etc. The new present tense form wol, with a rounded vowel, the basis of the contemporary negative won’t, originated in the Midlands as a back-formation from the preterit wolde.
4 Remarks on word formation As in the earlier periods, ME word formation involved processes such as compounding, prefixation, suffixation, derivation by a zero-morpheme, and back-derivation, the first three supplying the majority of new lexical items. ME writers continued to create new compounds but not as extensively as Anglo-Saxon anonymous poets, which is testified by 1069 compounds in the poem Beowulf compared with 800 compounds in Laȝamon’s Brut, a text five times longer than the former, cf. Burnley (1992: 441) and Sauer (1988). One of the reasons for that decline must have been the influence of French and Latin, from which numerous affixes were borrowed. Some of them became productive in English.
4.1 Compounding The OE technique of creating new words by compounding, i.e. putting two roots together, continued to be employed in Middle English. Sometimes it is difficult to evaluate whether we are dealing with a compound or with a “syntactic group”, e.g. dai-liȝt, daies liȝt and liȝt of daie (Sauer 1988: 187; cf. also 1992). Although less numerous than in Old English, ME compounds represented new types, for instance, with a verb in initial
27 Middle English: Morphology position (leep-yeer ‘leap-year’) or sex-determining compound nouns, with a pronoun in initial position, e.g. he-lamb, she-ass, she-ape (cf. Marchand 1969: 75–79). Such compounds represented both endocentric (goggle-eye) and exocentric structures (barefoot). Also predicate-object compounds in -ing formed a new category of words, e.g. backbiting, grisbitting, together with adjectives representing the newborn type or with locative particles like over, attached initially, which “come to be used figuratively in compounds (as overking, overlord, from about 1200)”; cf. Strang (1970: 259–260). The period also witnessed the growing popularity of compound agent nouns like man-slayer, gooddoer, soothsayer, peacemaker (14th–15th centuries). The same period saw the rise of compound adjectives, like icy-cold or lukewarm. Simultaneously, verbal compounds were often matched by particle verbs, which indicated a drift towards analytic structures, e.g. outflee and flee out, overlook and look over, etc. (Burnley 1992: 445). In some cases French loanwords replaced English compounds, cf. chueteine ‘chieftain’ substituting for OE heretoga ‘consul, army chief’, abbey for OE munucclif, or marbre ‘marble’ for OE marmon-stane.
4.2 Derivation Certain prefixes, like a-, be-, for-, to-, un- and others continued being attached in Middle English, together with the unproductive continuation of ȝe-, surviving as i- especially in the South. These were joined by loan-prefixes like de-, dis-, en-, in-, mis-, re-, sub-, super-, but their vast majority were unproductive. In Early New English, some of them became productive (en-, non-, etc.). Derivation by suffixing was one of the principal methods of creating new words. Especially frequent were the adjectival suffixes -ful, -ish as well as -ling found in nouns (cf. Burnley 1992: 448). More than twenty suffixes surviving in Middle English were joined by numerous suffixes of French origin. Here belong such common items as -able, -a¯ge, -al, -a¯cioun, -erie, -ess, -itee, -ment, -ous, etc. While suffixes like -a¯ge, -ity, and -ment are only attached to Romance bases, -able can be found after both native and foreign stems. As regards the foreign morphological element in English, all major controversies concern the question of whether French suffixes were productive in Middle English (cf. the standard studies by Koziol 1937; Jespersen 1942; Marchand 1969). Based on her Helsinki Corpus data, Dalton-Puffer (1996) answers the question negatively, whereas Miller (1997) claims that they became productive in Late Middle English. Of special importance is a pair of papers by Kastovsky (e.g. 1982, 1996). For a state-of-the-art report, cf. Ciszek (2008: 21–30, 110). For a relatively comprehensive listing of examples of different word-formation types in Middle English, see Fisiak (1996: 104–112).
5 Major problems in Middle English inflectional morphology Unlike ME phonological processes, such as Open Syllable Lengthening or the Great Vowel Shift, whose implementation has raised numerous interpretative controversies, the chronologically parallel morphological developments have failed to trigger similar heated disputes. Morphological changes in Middle English were characterized by a drift towards reducing inflections and transforming synthetic into analytic structures. The general accounts of the morphological processes are Moore (1928), O’Neill (1980,
431
432
IV Middle English 1982), Kastovsky (1988 on the typology of morphological change; 2007), and the monographs on English historical morphology by Bammesberger (1975), Wolff (1975), Faiss (1992), and Wełna (1996). As regards developments in inflection, they affected practically all parts of speech. Thus, apart from the emergence of the definite article contrasting with the indefinite one, Middle English suffered the loss of the dative and accusative cases. More important are, however, two other developments: the first is loss of the PL -en and other types of plural marking, which coincided with the spread of the PL -es; cf. Roedler (1911, 1916), Peters (1985), and Newman’s recent study (2008). The second is the substitution of semantic for grammatical gender (cf. Moore 1921; Jones 1988; Markus 1988) which was the consequence of the reduction of gender distinctive endings in nouns. The most important development in adjectives was, apart from the loss of inflections, the initial stage of the replacement of synthetic comparatives and superlatives by periphrastic forms with more and most, a process which began at the close of the 15th century; cf. Pound’s (1901) traditional account and the recent studies by Janecka (2008) and Gonza´les-Dı´az (2008: Chapter 3). As to the system of personal pronouns, discussions concentrate around the pronoun she whose appearance was due to functional reasons (cf. Britton 1991) following the merger of the feminine he¯o with masculine he¯. Similar factors determined the borrowing of the set of 3P PL pronouns they, their, them from Scandinavian (Stenroos 2005). Also verbs show loss of inflections, although not as intensively as the nouns. Some endings were replaced, as was the indicative PRES PL marker -eth (from OE -aþ) by -en, while 3P PRES SG -eth continued to be used. The origin of -ing, either marking the PRESPRTC (Rooth 1941/1942, Gleissner 1979) or attached to mark the gerund ( Jack 1988) still remains a controversial issue (cf. Budna 2007). Several studies deal with the decay of PASTPRTC prefixal marking; e.g. Pilch (1955a, b), and a recent study by Wojtys´ (2008). Perhaps the most spectacular event in the area of verbs was connected with the disintegration of the ablaut system whose effect was the shift of strong (ablaut) verbs to the class of weak verbs. How this complicated process operated in Middle English is discussed directly in Michelau (1910), Wełna (1990), Krygier (1994; the best study of the problem to date), Kahlas-Tarka (2000), and indirectly in Long (1944; focused on late ME). An account of suppletion in ME morphology (e.g. bad, went, syndon, be) is Hogg (2003).
6 References Bammesberger, Alfred. 1975. A Sketch of Diachronic English Morphology. Regensburg: Pustet. Britton, Derek. 1991. On Middle English she, sho: A Scots solution to an English problem. NorthWest European Language Evolution 17: 3–51. Brunner, Karl. 1962. Die englische Sprache. Vol. II. Tu¨bingen: Niemeyer. Brunner, Karl. 1963. An Outline of Middle English Grammar. G. K. W. Johnston (trans.). Oxford: Blackwell. Budna, Anna. 2007. On the origin of the present participle marking in mediaeval English. In: A. Weselin´ski and J. Wełna (eds.), Explorations in Literature and Language, 111–126. Warszawa: Wydawnictwa Uniwersytetu Warszawskiego.
27 Middle English: Morphology Burnley, David. 1992. Lexis and semantics. In: Norman Blake (ed.), The Cambridge History of the English Language, Vol. II. 1066-1476, 409–499. Cambridge: Cambridge University Press. Ciszek, Ewa. 2008. Word Derivation in Early Middle English. Frankfurt am Main: Peter Lang. Clark, Cecily. 1957. Gender in “The Peterborough Chronicle”. English Studies 38: 109–115. Dalton-Puffer, Christiane. 1996. The French Influence on Middle English Morphology: A CorpusBased Study of Derivation. Berlin/New York: Mouton de Gruyter. Faiss, Klaus. 1992. English Historical Morphology and Word-Formation: Loss Versus Enrichment. Trier: Wissenschaftlicher Verlag. Fisiak, Jacek. 1996. A Short Grammar of Middle English. Warszawa: Wydawnictwo Naukowe PWN. Fries, Charles Carpenter. 1938. Some notes on the inflected genitive in Present-Day English. Language 14: 121–133. Gleissner, Reinhard. 1979. Middle English -ind>-ing? and Bavarian -ind>-ing: A note. In: O. Hietsch (ed.), Bavarica Anglica, Vol. I. A Cross-Cultural Miscellany Presented to Tom Fletcher, 53–60. Frankfurt am Main: Peter Lang. Gonza´lez-Dı´az, Victorina. 2008. English Adjective Comparison: A Historical Perspective. Amsterdam/ Philadelphia: John Benjamins. Graband, Gerhard. 1965. Die Entwicklung der fru¨hneuenglischen Nominalflexion. Dargestellt vornehmlich auf Grund von Grammatikerzeugnissen des 17. Jahrhunderts. Tu¨bingen: Niemeyer. Hogg, Richard M. 1992. Phonology and morphology. In: Richard M. Hogg (ed.), The Cambridge History of the English Language, Vol. I. The Beginnings to 1066, 67–167. Cambridge: Cambridge University Press. Hogg, Richard M. 2003. Regular suppletion. In: Raymond Hickey (ed.), Motives for Language Change, 71–81. Cambridge: Cambridge University Press. Jack, George. 1988. The origins of English gerund. North-Western European Language Evolution 12: 15–75. Janecka, Joanna M. 2008. Periphrastic and Suffixal Adjectival Grading in Middle English (12th– 14th c.). Warsaw: Institute of English Studies, University of Warsaw. Jespersen, Otto. 1942. A Modern English Grammar on Historical Principles, Vol. VI. Morphology. London: Allen & Unwin, Ltd. Jones, Charles. 1988. Grammatical Gender in English: 950 to 1250. London: Croom Helm. Kahlas-Tarka, Leena. 2000. A note on non-standard uses in Middle English: Weak preterits of strong Old English verbs. Neuphilologische Mitteilungen 101: 217–223. Kastovsky, Dieter. 1988. Typological changes in the history of English morphology. In: Udo Fries and Martin Heusser (eds.), Meaning and Beyond. Ernst Leisi zum 70. Geburtstag, 160–178. Tu¨bingen: Gunter Narr. Kastovsky, Dieter. 1996. Morphological reclassification: The morphological and morphophonemic restructuring of the weak verbs in Old and Middle English. In: J. Klein and D. Vanderbeke (eds.), Anglistentag 1995 Greifswald. Proceedings 17, 273–284. Tu¨bingen: Niemeyer. Kastovsky, Dieter. 2007. Middle English word-formation: A list of desiderata. In: Gabriella Mazzon (ed.), Studies in Middle English Forms and Meanings, 41–56. Frankfurt am Main: Peter Lang. Koziol, Herbert. 1937. Handbuch der englischen Wortbildungslehre. Heidelberg: C. Winter. Krygier, Marcin. 1994. The Disintegration of the English Strong Verb System. Frankfurt am Main: Peter Lang. Long, Mary McDonald 1944. The English Strong Verb from Chaucer to Caxton. Menasha, WI: Banta. Marchand, Hans. 1969. The Categories and Types of Present-Day English Word-Formation. 2nd edn. Mu¨nchen: Beck. Markus, Manfred. 1988. Reasons for the loss of gender in English. In: Dieter Kastovsky and Gero Bauer (eds.), Luick Revisited, 241–258. Tu¨bingen: Narr. ¨ bertritt starker Verba in die schwache Coniugation. Ko¨nigsberg: Karg Michelau, Erich. 1910. Der U und Manneck.
433
434
IV Middle English Miller, Gary. 1997. The morphological legacy of French: Borrowed suffixes on native bases in Middle English. Diachronica 14: 233–264. Minkova, Donka. 1990. Adjectival inflection relics and speech rhythm in Late Middle and Early Modern English. In: Sylvia Adamson, Vivien A. Law, Nigel Vincent, and Susan Wright (eds.), Papers from the 5th International Conference on English Historical Linguistics, Cambridge, 6–9 April 1987, 313–336. Amsterdam/Philadelphia: John Benjamins. Moore, Samuel. 1921. Grammatical and natural gender in Middle English. Publications of the Modern Language Association 36: 79–103. Moore, Samuel. 1928. Earliest morphological changes in Middle English. Language 4: 238–266. Newman, John G. 2008. The Spread of the s-Plural Formative in Old and Middle English Nouns. Warsaw: Institute of English Studies, University of Warsaw. O’Neil, Wayne. 1980. The evolution of the Germanic inflectional systems: A study in the causes of language change. Orbis 27: 248–286. Peters, Robert A. 1985. Historical development of noun plural -(e)s. Journal of English Linguistics 18: 25–32. Pilch, Herbert. 1955a. Der Untergang des Pra¨verbs ge- im Englischen. Anglia 73: 37–64. Pilch, Herbert. 1955b. ME. i- beim Participium Pra¨teriti. Anglia 7: 279–291. Pound, Louise. 1901. The Comparison of Adjectives in English in the XV and the XVI Century. Heidelberg: Carl Winter’s Universita¨tsbuchhandlung. Roedler, Eduard. 1911. Die Ausbreitung des s-Plurals im Englischen (1). Ph.D. dissertation, Kiel. Roedler, Eduard. 1916. Die Ausbreitung des s-plurals im Englischen (2). Anglia 4: 420–502. Rooth, Erik. 1941/1942. Zur Geschichte der englischen Partizip-Pra¨sens-Form auf -ing. Studia Neophilologica 1: 71–85. Sauer, Hans. 1988. Compounds and compounding in Early Middle English: Problems, patterns, productivity. In: Manfred Markus (ed.), Historical English. On the Occasion of Karl Brunner’s 100th Birthday, 186–209. Innsbruck: AMOE. Sauer, Hans. 1992. Nominalkomposita im Fru¨hmittelenglischen. Mit Ausblicken auf die Geschichte der englischen Nominalkomposition. Tu¨bingen: Niemeyer. Stenroos, Merja. 2005. The spread of they, their and them in English: The Late Middle English evidence. In: Marcin Krygier and Liliana Sikorska (eds.), Naked Words in English, Frankfurt am Main: Peter Lang. Strang, Barbara M. H. 1970. A History of English. London: Methuen. Wełna, Jerzy. 1990. The strong-to-weak shift in English verbs: A reassessment. Kalbotyra 42(3): 129–139. Wełna, Jerzy. 1996. English Historical Morphology. Warszawa: Wydawnictwa Uniwersytetu Warszawskiego. Wełna, Jerzy. 2001. Suppletion for suppletion, or the replacement of e¯ode by went in English. Studia Anglica Posnaniensia 36: 95–110. Wełna, Jerzy. 2005. Nim or take? A competition between two high frequency verbs in Middle English. Studia Anglica Posnaniensia 41: 53–69. Wojtys´, Anna. 2008. Past Participle Marking in Mediaeval English: A Corpus-Based Study in Historical Morphology. Warsaw: Institute of English Studies, University of Warsaw. Wolff, Dieter. 1975. Grundzu¨ge der diachronischen Morphologie des Englischen. Tu¨bingen: Niemayer. Wright, Joseph and Elizabeth Mary Wright. 1928. An Elementary Middle English Grammar. London: Oxford University Press.
Jerzy Wełna, Warsaw (Poland)
28 Middle English: Syntax
435
28 Middle English: Syntax 1 2 3 4 5
Locating syntax The syntax of Middle English prose The syntax of Middle English verse Form and function References
Abstract Recent research has emphasized the importance of performance/parole as much as competence/langue in linguistic theory; a good example is William Kretzschmar’s (2009) The Linguistics of Speech. This article offers a qualitative analysis of a series of four short Middle English texts, both in verse and prose, to illustrate developments in syntax during the period. These analyses engage not only with modern notions of syntactic structure but also with medieval ideas; whereas many modern grammarians reify the grammatical conception of the sentence, medieval thinking focused on rhetorical structure, in which the sentence was primarily a semantic notion and the units of analysis were the “period”, the “colon” and the “comma”. Issues of variation – dialectal, diachronic and genre-driven – are also addressed. Topics arising from these analyses and discussed here include coordination and subordination, element-order practices, concord, changes in the structure of noun- and verb-phrases, and negation.
1 Locating syntax Linguists traditionally distinguish between two axes of grammatical analysis: “paradigmatic” and “syntagmatic”. The paradigmatic axis deals with the relationship between forms of lexemes, e.g. love, loves, loved or book, books, book’s, while the syntagmatic axis deals with the relationships between lexemes, e.g. the relationships between subject, predicator and object in a clause such as She loves him. The paradigmatic axis is often referred to as inflectional morphology. The syntagmatic axis engages with the study of syntax. To put it another way, the paradigmatic axis focuses on linguistic form, while the syntagmatic axis engages primarily with linguistic function. (Lexical morphology, often referred to as word-formation, is not discussed here. Word-formation deals with the creation of new lexemes through the linking of morphemes, e.g. blackbird from black and bird, or disgraceful from combining grace with dis- and -ful. Historically, there are also interesting questions as to whether morphemes were seen as forming one or more lexemes; the common practice in Early Modern English of printing shalbe ‘shall/must be’ as one word rather than two suggests that these two words were conceived of as forming one unit. Similar issues apply in ME; thus, for instance, in example (1) below the item ‘upon’ appears as up on in the manuscript, while in example (4) ‘amiss’ appears as a mys. Word-formation has clearly both paradigmatic and syntagmatic aspects, but on grounds of space and clarity it is not discussed further here; see further Wełna, Chapter 27, Section 4.) Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 435–450
436
IV Middle English Until fairly recently it was commonplace to emphasize the discrete character of linguistic categories, but recent work – influenced by developments in cognitive studies – has drawn attention to the fuzzy divisions between these traditional distinctions: whether a particular construction is discussed under the heading of lexical morphology or syntax depends on the definition of the notion “word”, while of course paradigmatic selection in a clause such as She loves him depends on the syntagmatic relations which obtain between the chosen forms. There is therefore a difficulty in separating the study of syntax from the investigation of other aspects of grammar – or indeed in distinguishing grammar from the study of the lexicon (lexicology), since grammar and the lexicon are both ways whereby meaning (semantics) is instantiated, and thence transmitted through speech or writing. It is not unexpected that much recent research in historical syntax has engaged with issues such as grammaticalization, i.e. the process whereby the instantiation of meaning is transferred from lexical to grammatical expression (see further Hoffmann and Brems, Volume 2, Chapter 99). Any discussion of ME syntax, therefore, needs to engage not only with syntax as traditionally delimited but also with the overlaps which obtain between syntax and other aspects of linguistic enquiry. A second major development in the study of syntax is in some sense operational, but has a theoretical implication. Developments in corpus studies, encouraged by groundbreaking work on historical data at the University of Helsinki in particular, have led to a new interest in evidential matters (see Kyto¨, Volume 2, Chapter 96). It is no coincidence that there has been a consequent shift in theoretical studies, from a focus on competence/langue to a new interest in performance/parole. The study of electronic corpora allows researchers to observe the emergence of new structures both diachronically (i.e. through time) and diatopically (i.e. through space); it also enables a reassessment of the traditional concept of well-formedness, whereby researchers used their native-speaker intuition to determine whether or not a particular sentence was acceptable or not. Native-speaker intuition is of course not available to historical linguists, but corpora supply at least some data, enabling us to identify dominant patterns which may address that lacuna. A focus on parole rather than langue reminds researchers that notions such as “sentence” are, like “word”, less straightforward than have been often assumed. Medieval grammarians considered the sententia to be primarily a semantic rather than a grammatical notion while the primary unit for analysis was the periodus or period, i.e. “[…] an utterance or complete rhetorical structure which expresses a single idea or sententia […]” (Parkes 1992: 306). The sentence was thus a “thought or opinion; especially the substance or significance expressed by the words of […] a rhetorical ‘period’ ” (Parkes 1992: 307). Medieval rhetoricians also distinguished divisions within the period: the colon (PL cola) and, within the colon, the comma (PL commata) (these terms were later transferred to the punctuation-marks which evolved originally to distinguish them). These divisions were traditionally flagged in speech through the use of rhythmical features where it was necessary to pause to a greater or lesser extent. They are therefore essentially rhetorical units, correlating with patterns of speech and employed as guidance to assist reading aloud; written texts are seen as secondary aids to the primary method of communication, viz. speech. In the medieval period, the comma was often marked by the virgula suspensiva or virgule , or sometimes the punctus or point , while the colon was marked typically by the punctus elevatus יִthe period
28 Middle English: Syntax was marked by a punctus but sometimes by the use of litterae notabiliores (“more notable letters”, i.e. capitals). Such notions not only underpin the very various punctuationpractices of ME scribes but they also, of course, challenge universalist claims in modern linguistic categorization. An attempt will be made here to harness these new insights, in a preliminary way, to studying Middle English (hence ME) syntax. The main resources to be used other than standard editions will be the most recent version of the Stavanger-Glasgow Middle English Grammar Corpus (MEG-C) (Stenroos et al. 2011), and the corpus developed for the Edinburgh Linguistic Atlas of Early Middle English (LAEME) (Laing and Lass 2007). Discussion will be informed by illustration: short texts in verse and prose will be presented and analyzed. These texts will be presented in what may seem to some readers (and users of other corpora) an unfamiliar guise: they will appear in “diplomatic” editions, whereby an attempt is made to reflect not only the spellings but also the punctuation (or lack of punctuation) of the original. Such editions engage with the text as it is, rather than present it through the interpretative lens of modern critical editors whose concerns often skew our reading anachronistically. Moreover, these texts will be chosen from different centuries within the ME period and from different geographical areas. It should be noted that a chapter on “ME syntax” should not reify a single kind (let alone a “standard” variety) of Middle English. For various reasons discussed elsewhere (see Schaefer, Chapter 33; Lange, Chapter 62), Middle English was not only, like Present-day English (PDE), a congeries of varieties, but these varieties were reflected in the written as well as in the spoken mode. A distinction is made between prose and verse, though that distinction was arguably looser in ME times than it is today. Such a short chapter cannot of course supply a comprehensive outline of ME syntax; the distinction made here between verse and prose, for instance, is undoubtedly crude, and generic considerations demand a much more nuanced approach than is possible here (see Arnovick, Chapter 35). This chapter, which focuses on qualitative rather than quantitative approaches, is intended simply as a starting-point for discussion.
2 The syntax of Middle English prose MS Cambridge, Corpus Christi College 402 is generally dated to the second quarter of the 13th century, perhaps about 1230 (see Laing 1993: 24 and references there cited). The text’s language has been localized to Herefordshire or Shropshire (see Laing 1993: 24, although see also Millett 2005 and subsequently Laing and Lass 2007). The manuscript contains a single item: the Ancrene Wisse (‘Guide for Anchorites’). The Corpus manuscript of Ancrene Wisse has been the basis for most modern editions of this important piece of ME religious prose, ever since the pioneering work of J. R. R. Tolkien (1929; see also Shepherd 1972; Millett 2005). It represents a variety of ME which was both conservative and innovatory. Copies of the Old English (hence OE) homilies of Ælfric and Wulfstan were still being produced in the south-west Midlands when the work was composed, and there are aspects of this text which demonstrate the impact of the Anglo-Saxon prose-tradition; but the author was also clearly well-acquainted with issues current in contemporary Parisian intellectual circles, and was therefore in touch with the French-centered 12th-century renaissance. The form
437
438
IV Middle English of ME prose which resulted, devised originally for a “gentle and lettered” (D’Ardenne 1961: 177) female audience/readership, reflects these two influences. The following passage, (1), from Part VII of the text and based on Tolkien’s diplomatic edition (Tolkien 1962: 198–199), is often quoted in studies of the text, although generally in a form different from how it appears in the manuscript. The passage allegorizes the soul as a lady besieged in a castle, offered the support of a mihti kinges luue ‘a powerful king’s love’; this king is later revealed to be iesu godes sune ‘Jesus, God’s son’. Expanded abbreviations are underlined. (1)
A leafdi wes mid hire fan יִbis et al abuten. hire lond al destruet. & heo al poure inwið an eorðene castel. A mihti kinges luue wes þah biturnd up on hire swa unimete swiðe יִþet he for wohlech sende hire his sonden. an efter oðer. ofte somet monie. sende hire beawbelez baðe feole & feire. sucurs of liueneð. help of his hehe hird to halden hire castel. Heo underfeng al as on unrecheles. & swa wes heard iheortet יִþet hire luue ne mahte he neauer beo þe neorre. hwet wult tu mare he com him seolf on ende. schawde hire his feire neb. as þe þe wes of alle men יִfeherest to bihalden. spec se swiðe swoteliche. & wordes se murie יִþet ha mahten deade arearen to liue. wrahte feole wundres & dude muchele meistries biuoren hire ehsihðe. schawde hire his mihte. talde hire of his kinedom. bead to makien hire cwen of al þet he ahte. al þis ne heold nawt. nes þis hoker wunder? for heo nes neauer wurðe forte beon his þuften. ah swa þurh his deboneirte luue hefde ouercumen him יִþet he seide on ende. Dame þu art iweorret. & þine van beoð se stronge יִþet tu ne maht nanesweis wið ute mi sucurs edfleon hare honden. þet ha ne don þe to scheome deað efter al þi weane. Ich chulle for þe luue of þe יִneome þet feht up o me. & arudde þe of ham þe þi deað secheð. Ich wat þah to soðe þet ich schal bituhen ham neomen deaðes wunde. & ich hit wulle heorteliche forte ofgan þin heorte. Nu þenne biseche ich þe for þe luue þet ich cuðe þe. þet tu luuie me lanhure efter þe ilke dede dead יִhwen þu naldest liues. þes king dude al þus. arudde hire of alle hire van. & wes him seolf to wundre ituket & islein on ende. þurh miracle aras þah from deaðe to liue. Nere þeos ilke leafdi of uueles cunnes cunde. ʒef ha ouer alle þing ne luuede him her efter? (Ancrene Wisse; Tolkien [ed.] 1962: 198–199) ‘A lady was surrounded all about by her enemies, her land entirely destroyed, and she entirely poor, inside an earthen castle. A powerful king’s love was nevertheless directed at her, so very great that he, in order to woo her, sent to her his messengers, one after another, often many together. He sent her jewels, both many and beautiful, succour in materials for existence, help from his noble court to defend her castle. She received everything in a careless fashion, and was so hard-hearted that he could never be any nearer her love. What more do you want? He came himself at last, showed her his handsome face which was, in comparison with all men, most beautiful to look at; he spoke so very sweetly, and words so pleasant that they could raise the dead to life; he performed many marvels and did many exceptional deeds in front before her eyes; he showed her his power, told her of his kingdom, offered to make her queen of everything that he owned. All this had no effect. Was this disdain not a marvel? – for she was never worthy to be his maidservant. But even so through his generosity love had overcome him, so that he said at last: “Lady, you are attacked, and your enemies
28 Middle English: Syntax are so strong that you could not in any way escape from their hands without my help, lest they put you to a shameful death after all your sorrows. I must because of my love for you take that battle upon myself, and rid you of those who seek your death. I know nevertheless in truth that I must receive death’s wound from them, and I wish it heartily in order to win your heart. Now then I beseech you, in the name of the love I feel for you, that you may love me when I am dead after this deed, when you would not so while I was alive”. This king did everything thus; he rid her of all her enemies, and was himself terribly tormented and eventually killed. Through a miracle he arose nevertheless from death to life. Would not this same lady have an evil-natured character is she did not love him thereafter above all things?’ (For the convenience of readers, translations into Present-day English of all the passages in this chapter are offered here. It will be observed that PDE punctuation has been supplied, although there are of course distortions which necessarily follow. For a modern translation of Ancrene Wisse, see White 1993.) Four marks of punctuation are used in this passage: the punctus, the punctus elevatus, the punctus interrogativus , and litterae notabiliores. The punctus and punctus elevatus are used to distinguish major and minor pauses respectively, the punctus interrogativus is used to indicate a question, while litterae notabiliores are used sparingly to mark major shifts in the rhetoric of the passage; thus A mihti kinges luue contrasts with the lady’s response, flagged by the capitalized Heo ‘she’; and stages in the king’s statements are marked by Dame, Ich and Nu. The syntax of the passage is highly complex. The following section demonstrates the impact of two stylistic modes, crudely distinguished as (a) parataxis, and (b) hypotaxis: hwet wult tu mare he com him seolf on ende. schawde hire his feire neb. as þe þe wes of alle men יִfeherest to bihalden. spec se swiðe swoteliche. & wordes se murie יִþet ha mahten deade arearen to liue. wrahte feole wundres & dude muchele meistries biuoren hire ehsihðe. schawde hire his mihte. talde hire of his kinedom. bead to makien hire cwen of al þet he ahte.
In parataxis, units are placed in parallel with each other and readers/audiences are invited to make connexions themselves, while in hypotaxis hierarchical relationships between units are made explicit: we might compare “I came, I saw, I conquered” (parataxis) with “I conquered because I saw. Because I came, I saw” (hypotaxis). Periods such as schawde hire his feire neb, spec se swiðe swoteliche, talde hire of his kinedom are deployed paratactically; close examination of the passage shows that the section includes a series of verbs all dependent on he: schawde, spec, wrahte, dude, schawde, talde and bead. But interposed within these paratactic elements are modifying/subordinated constructions: as þe þe wes of alle men יִfeherest to bihalden, and þet ha mahten deade arearen to liue. The technique is an example of the author’s “careful involved symmetry” (Shepherd 1972: lxxii), frequently demonstrated elsewhere in the work: “symmetry, parallelism, and antithesis are habits of his thought” (Shepherd 1972: lxvii). This kind of sophisticated usage clearly derives in part from the models supplied by the great homilists of the Anglo-Saxon period, but it also shows the impact of the rhetorical arts current in France, most notably the preaching-style developed by writers such as Maurice of Sully.
439
440
IV Middle English It is interesting that the rhetorical question hwet wult tu mare which opens this sequence is not distinguished by a punctus interrogativus or even a punctus or punctus elevatus; the passage moves swiftly to the sequence of statements which follow, and no pause is suggested by the Corpus scribe. Other important early manuscripts of the text, roughly contemporary with the Corpus MS, manifest different behaviors. Whereas the principal scribe of MS London, British Library Cotton Cleopatra C.vi shares the usage of the Corpus scribe, the scribe of MS London, British Library Cotton Nero A. xiv places a punctus after the equivalent period, thus, hwat wult tu more. (It should be noted, however, that the Nero scribe tends to use punctuation more insistently than the Corpus and Cleopatra scribes. Parallel passages from the various texts may be examined in Day 1952: 177, Dobson 1972: 284–285.) It is possible to distinguish dominant element-order patterns in example (1). The commonest usage is the ordering subject-verb, as in Heo underfeng al […], he com him seolf on ende. However, constructions such as A leafdi wes mid hire fan יִbis et al. abuten, where the interposed prepositional phrase mid hire fan and punctus elevatus suggests that the past participle biset is still perceived as a complement rather than part of the verb phrase, indicate that the PDE complex verb phrase has still not wholly emerged. As was frequently the case in Old English (and as is still the case in Present-day German), an initial adverbial construction causes the position of the following subject and verb to be reversed in sequence, e.g. Nu þenne biseche ich þe. In subordinated periods, more flexible element-orders are possible, e.g. þet hire luue ne mahte he neauer beo þe neorre, and in questions verbs precede subjects, e.g. nes þis hoker wunder?, Nere þeos ilke leafdi of uueles cunnes cunde [….]?. A delayed verb appears in ʒef ha ouer alle þing ne luuede him her efter, while we might note the verb at the end of the period in the relative construction þe þi deað secheð. Double negation is used, apparently for emphatic reasons: al þis ne heold nawt (we might note in this last example the intransitive use of heold, cf. OED hold, vb. sense 25) beside single negation in ha ouer alle þing ne luuede him; constructions with ne preceding the verb are common in Early ME prose, shifting to not – predominantly but by no means exclusively post-verbal – later in the ME period. Contracted forms are also commonly used, e.g. Nere, naldest etc; such forms seem to have been innovative in late OE but recessive as the ME period progressed (see for all these findings, and for further discussion of negation, Iyeiri 2001). Within noun phrases agreement of adjective and noun is sustained, e.g., muchele meistries, with an adjective inflected in -e, or in the genitive phrase within the prepositional construction of uueles cunnes cunde. Morphological change has clearly had an impact on syntactic usage; inflectional endings in this passage, though of a wider variety than in Present-day English, are clearly reduced in comparison with Old English, and this development has syntactic implications. Thus both forms in the noun phrase feole wundres ‘many marvels’ derive from Old English, but OE fela ‘many’, which seems to have been classified as a numeral, was followed by the plural genitive, thus OE fela wundra. The emerging indefinite article appears in an eorðene castel; Old English, which had a two-way deictic system (“simple” versus “compound”), did not have an equivalent to this construction (see e.g. Jones 1988). Prepositional use is more widespread in example (1) than was the case in Old English. Verb phrases similarly reflect certain OE patterns but modified to reflect linguistic developments. We might note the retention of the formal subjunctive in certain
28 Middle English: Syntax subordinate clauses, e.g. þet tu luuie me lanhure efter þe ilke dede dead (cf. indicative luuest; D’Ardenne 1961: 234) or in questions, e.g. Nere þeos ilke leafdi of uueles cunnes cunde …?; auxiliaries such as maht(en) seem not yet to have taken over the subjunctive function they have in Present-day English, although of course there is a semantic overlap (see Smith 1996a: 151–153). Similarly, forms such as schal, chulle, naldest, wulle, etc. have not yet taken over fully the role of future auxiliary, and are still used with their “full” lexical meaning, i.e. ‘must’ for schal etc. and ‘wish’ for wulle etc. Lexical uses of these words in example (1) are clear in & ich hit wulle heorteliche forte ofgan þin heorte, if less so in Ich chulle for þe luue of þe יִneome þet feht up o me; the process of grammaticalization of these forms seems incipient rather than complete (see further Warner 1992; see also Denison 1993: 292–339). A contrasting passage (example 2) in a very different genre comes from the opposite end of the country and from later in the Middle English period: a copy of the so-called “defective” version of the prose Mandeville’s Travels, once MS Cambridge, Bradfer Lawrence 7 and now MS Tokyo, Takamiya 63 (see Seymour 2002). The language of the text has been localized by the Linguistic Atlas of Late Mediaeval English (McIntosh et al. 1986) to the West Riding of Yorkshire, as LP 1349; the manuscript dates from the first quarter of the 15th century. The passage is taken (with some modifications and the omission of some lines, marked […]) from the first tranche of the text printed in MEG-C (Stenroos et al. 2011). The passage printed here corresponds to chapter 5 and the beginning of chapter 6 of the work in Seymour’s critical edition (2002: 21–23). As in example (1), expanded abbreviations are underlined. (As is common in many northern varieties of Middle English, the letter “thorn” is represented in example (2) by the figura , e.g. as in yat ‘that’, and is thus written in a way indistinguishable from “y”. Not all northern varieties, however, collapse the distinction, as will be noted in example (4) below. See further Benskin 1982. For the interpretation of the place-names in example (2), see Seymour 2002.) (2)
And who so wille ga to ye land of babilone יִwher ye sawtene duelles to haue leue to ga ye mar sykerle yoro ye contrees יִand for to ga to mounte Synay be for he come to jerusalem. and yan turne a gayne to jerusalem he sall ga fra gasa to ye cas[t]elle dayre and aftur men comes oute of s[u]rry & gas in to wildernese. wher ye way is fule sandy and yatt wildernese lastes. viij iournese whar men fyndes alle att yam neddyse of vitayles יִand men calles yat wildernese of achelleke & יִwhen a man comes out of desertte he entres in to Egipte canapote & יִin a noyer langage men calles itt Mercyne יִand ye firste goode toune yat men fyndes is called beley and itt is of ye kyngdome of alape יִAnd fro yennes men comes to babilone. and yer is a fayr kyrke of our ladye. wher scho duellyde. vij ʒeer when scho wentte. when scho went [dittography] oute of lande of Jnde יִffor drede of kynge heroude יִ and yar lygges ye body of saynte barbare. and yer duellyde on iosephe when he was saulde of his bredyr & יִyar made Nabogodonoser thre childre be brentte er yai wer in ryʒte trouthe. whilke childer men calles יִAnanya יִaʒarya יִMisael יִas ye psalme of benedicite telles יִBott nabogodonoser callede yame yus יִSidrake Misaake abdenego. yat is to say. gude glorius gode. vittorius gode. our alle kyngdome and yatt was for ye grett myracle yatt he sawe gode ga. yoro ye fyer with yes childer יִyar duelles ye. souden. ffor yar is a fayr see in a Castelle. and itt is strange. & well sette on a roche Jn yatt Castell. is duellynge to kepe yatt Castelle
441
442
IV Middle English & to ser fe ye sowdane יִmo yane viij thosande persones of folke יִyatt takes alle ye nessessaries of ye contre. of ye soudane J awghte well to witte יִFor J duellede with hym in his werres soudeour a grett whille. a gayne ye bedoynes יִand he walde haue weddyde me to a grett prynce douʒter ryghte Richelye ]…[ יִNow when aman has visetyde yis place of saynt katerine יִand he wille turne to Jerusalem יִhe sall firste take leefe att ye monkes. & Recomaunde hym specialie to yer prayere and ye same Monkes gifes with goode wille vitayles to pilgrymes to passe with yoro wildernese to Surry and yatt lastes well xiij iournese יִand in yatt Contre duelles many arabiens yat men calles bedoynese and ascoperdes. and yes er men of alle Euel condicoun & yai haue ne houses. bott tentes whilke yai make of Camelus & of oyer bestes yat yai ette and yar vnder yai lygge יִand yai duel. in places whar yai may fynde water of ye rede see ( יִMandeville’s Travels) ‘And whosoever wishes to go to the land of Cairo (babilone) where the sultan dwells, in order to get leave to go more safely through the countries and in order to go to Mount Sinai before he may come to Jerusalem, and then turn back to Jerusalem, he must go from Gaza to Castle Darum. And afterwards men come out of Syria and go into the wilderness, where the way is very sandy, and that wilderness lasts eight days’ journeying, where men take all that they need in provisions. And men call that the wilderness of Achelleke. And when a man comes out of the desert he enters into Egypt Canopus, and in another language men call it Mesryn. And the first good town that men find is called Beleth, and it is part of the kingdom of Aleppo. And from thence men come to Cairo, and there is a fair church of Our Lady, where she lived for seven years when she went from the land of Judaea for fear of King Herod. And there lies the body of Saint Barbara, and there Joseph continued to live when he was sold by his brothers. There Nebuchadnezzar caused three children to be burned before they were in the true faith, which children men call Ananiah, Azariah, Misael, as the psalm of blessing tells. But Nebuchadnezzar called them thus: Shadrach, Meschach and Abednego, that is to say, good God, victorious God, our entire Kingdom. And that was because of the great miracle that he saw God go through the fire with these children. The sultan dwells there. For there is a handsome seat in a castle, and it is strong, and well set on a rock. In that castle, to protect it and to serve the sultan, more than 8000 persons from that people are dwelling. I ought well to know, for I dwelt with him for a great while as a soldier in his wars against the Bedouins, and he wished to wed me very nobly to the daughter of a great prince […] Now when a man has visited this place of Saint Katherine, and he wishes to turn to Jerusalem, he must first take leave from the monks, and commend himself to their prayers. And the same monks give with good will provisions to pilgrims to pass with through wilderness to Syria, and that lasts a good thirteen days’ journey. And in that country dwell many Arabs whom men call Bedouins and Azoparts (i.e. Ethiopians). And these are men of an entirely evil condition, and they have no houses, except tents that they make from camels and other beasts that they eat, and under which they lie. And they dwell in places where they can find water from the Red Sea’ The main marks of punctuation in example (2) are the same as those in Ancrene Wisse, viz. the punctus and punctus elevatus (the passage does not include questions),
28 Middle English: Syntax accompanied by a few litterae notabiliores. These last are mostly used for proper names, but also sporadically in (2) for some conjunctions (And, Bott, ffor); the capital letter in the preposition Jn ‘in’ seems also to mark a new period. Here the punctus elevatus is much more common than the punctus, and seems to be used by the scribe to mark most periods as well as lesser divisions. The punctus seems to be used in this passage either to mark some subordinate structure or in lists, rather in the fashion of the present-day comma. However, as in PDE usage, some subordinated constructions are not so marked, e.g. the post-modifying unit att yam neddyse of vitayles. Examination of this passage shows a sustained dominance of parataxis: periods are commonly linked by and, &, etc. Hypotactic structures are much more sparingly used, e.g. whilke childer men calles […]. It is customary for historians of English prose style to detect a steady movement from parataxis to hypotaxis over time, correlating with increasing literacy in the vernacular, but the evidence from this passage is that the movement is not straightforward; indeed, in the 15th century, the paratactic style was still quite widespread. It is interesting, for instance, that Thomas Malory, when translating from French for his Arthurian cycle, chose to replace the hypotaxis of his French originals with a paratactic style which he clearly felt more appropriate (see Smith 1996b and references there cited). Element-order within periods is predominantly subject-verb, even in units beginning with a subordinating conjunction, e.g. when a man comes out of desertte. Verb-subject ordering after an adverbial, however, is retained in yar lygges ye body of saynte barbare, yar made Nabogodonoser thre childre be brentte […], in yatt Contre duelles many arabiens etc. Some variation is clearly still allowed, as in the relative construction whilke childer men calles […], but such usages are still found in formal Present-day English. The lengthy sequence Jn yatt Castell. is duellynge to kepe yatt Castelle & to ser fe ye sowdane יִmo yane viij thosande persones of folke יִmight also be noted, where the complex verb phrase is duellynge is positioned some way before the extended subject mo […] of folke. Noun-phrase syntax is essentially that of Present-day English. The -e inflection which was used in (e.g.) late 14th-century Southern Middle English to distinguish plurals and weak adjectives from strong singulars is no longer used in this way, as is to be expected in Northern Middle English at this date; thus ye grett myracle. Where -e is retained it seems to be simply a scribal flourish; thus whereas ye firste goode toune would seem to retain -e, we might also note gude glorius gode, where the form gude, being in “strong” position, would not have attracted an inflectional -e in (for instance) contemporary southern Chaucerian English. An uninflected genitive appears in a grett prynce douʒter; elsewhere in example (2) possession is commonly expressed by means of the preposition of, while we might also note the uninflected noun after a numeral in vij ʒeer. The indefinite article a is commonly deployed in the passage, e.g. a fayr see. Verb-phrase syntax is much closer to that used in Present-day English than was the case in example (1). In the period when aman has visetyde yis place of saynt katerine the past participle visetyde is not separated from has, as was habitual in example (1). A formal subjunctive, however, appears in be for he come to jerusalem, and may still best seems translated as ‘can’ in in places whar yai may fynde water of ye rede see יִ. The modals wille and sall (cf. PDE ‘shall’) still, moreover, seem to sustain lexical rather than grammatical force in and he wille turne to Jerusalem יִhe sall firste take leefe att ye monkes. (see further Warner 1982: 192–197). As is common in Northern Middle English,
443
444
IV Middle English plural subjects govern present-tense verbs in -es in men fyndes etc., and in ye same Monkes gifes with goode wille vitayles to pilgrymes. We might also note, however, yai make, yai ette, yai lygge, and yai duel, showing a distinct inflection when the verb is preceded by a pronoun. The text therefore illustrates the operation of the Northern Middle English/Older Scots Personal Pronoun Rule. Relics of this system still remain in certain rural varieties of Present-day Scots, and indeed elsewhere; its origin is much debated, but the current scholarly consensus is that its occurrence is the result of interaction with Celtic (see further Benskin 2011 and references there cited). The system works as follows: if the subject of the clause is a personal pronoun (i.e. ‘I’, ‘thou’, ‘he’, etc.), and comes immediately before or after the verb, the paradigm is as in Table 28.1: Table 28.1: Northern Middle English/Older Scots Personal Pronoun Rule SG
1 2 3
PL
I keip thou keipis he/scho/it keipis we/ʒe/thai keip
Otherwise the -is form is used throughout the paradigm. We might also note the verb phrase in Jn yatt Castell. is duellynge to kepe yatt Castelle & to serife ye sowdane יִmo yane viij thosande persones of folke יִ. The verb phrase is duellynge represents the appearance of the progressive present construction, rare in earlier English but subsequently commonplace (see further Brinton 1988). It will be observed therefore that example (2) displays both more “advanced” and more “conservative” features. The preference for parataxis might seem conservative, as might be the handling of verbs which later developed into auxiliaries; but (e.g.) the handling of element-order is, as might be expected from the later date, much more in line with PDE than was the case in example (1).
3 The syntax of Middle English verse MS London, British Library, Cotton Caligula A.ix (Part I), dated to the second quarter of the 13th century, has been localized, like the Corpus manuscript of Ancrene Wisse, to the south-west Midlands, although in this case to north-west Worcestershire (Laing 1993: 69–70; see also Laing and Lass 2007). It contains one of the two surviving texts of Lawman’s Brut, lines from which appear in example (3). Example (3) is based on Brook and Leslie’s edition (1963: 3–4), but corrected with reference to the text as it appears in LAEME (Laing and Lass 2007). Although verse, the passage is written out – as was the case with AngloSaxon poetry – as prose, and line-endings in the original manuscript are marked by │. However, example (3) is laid out below for the most part in pairs of “half-lines”, the basic metrical unit of Anglo-Saxon prosody, to reflect the verse structure of the text. (3)
Nu bidde[ð] laʒamon alcne│ æðele mon יִ for þene almiten godd. │ þet þeos boc rede & יִleornia þeos ru│nan. þat he þeos soðfeste word יִseg│ge to sumne. for his fader saule │יִþa hine for[ð] brouhte.
28 Middle English: Syntax & for his mo│der saule יִþa hine to monne iber. │ & for his awene saule יִþat hire│ þe selre beo. AmeN. NV seið mid loft songe þe│ wes on leoden preost. al│ swa þe boc spekeð יִþe he│ to bisne inom. Þa grickes hefden│ troye יִmid teone bi wonen. & þat lond│ iwest & יִþa leoden of slawen. & for│ þe wrake dome יִof Menelaus que│ne. elene was ihoten. alðeodisc│ wif. Þa paris alixandre יִmid pret│ wrenche. bi won. for hire weoren│ on ane daʒe יִhund þousunt deade. │ vt of þan fehte יִþe was feondli│che stor. Eneas þe duc יִmid ermðen│ at wond. Nefede he boten anne│ sune יִþe was mid him isund. │ Asscanius was ihoten יִnefede he│ bern no ma. & þes duc mid his drih│te יִto þare sæ him droh. of kunne│ & of folke יִþe fulede þan duke. of│ monne & of ahte. þe he to þare│ sæ brouhte. & tuenti gode scipen. │ he guðliche fulde. & þa scipen foren│ wide יִʒeon þare wintrede sæ.(Brut; Brook and Leslie [eds.] 1963) ‘Now Lawman [laʒamon] asks every noble man, for the sake of almighty God, who may read this book and discover these secrets, that he may recite appropriately this true speech, for the soul of his father, who brought him forth, and for the soul of his mother, who bore him as a man, and for his own soul, so that it may be safer. Amen. Now say with inspiration he who was priest among the people, just so the book speaks, which he took as an exemplary narrative. When the Greeks had tragically captured Troy, and laid the land waste, and slain the people, and for the avenging of Menelaus’s queen, who was called Helen, an alien woman (whom Paris Alexander won with practised treason), for her in a single day a hundred thousand died. Out of that fighting, which was extremely fierce, Aeneas the duke escaped with anguish. He had only one son who was saved with him, who was called Ascanius; he had no more children. And this duke retreated with his household to the sea, of family and of people who followed the duke, of men and of property which he brought to the sea. And he filled twenty good ships in martial fashion, and the ships went far and wide across the wintry sea’. (For an excellent up-to-date translation of the poem, see Allen 1992.) The punctuation of the passage is closely tied to its verse-structure. Thus the mid-line caesura is generally flagged by the punctus elevatus and the end of lines by a simple punctus; it will be observed that, in this passage at least, the occurrence of the caesura corresponds to a break between periodic units, such as phrases or clauses. Litterae notabiliores are deployed in general to mark steps in the argument, although they are also sporadically used for personal names. In general the syntax of the passage is comparable with that in example (1). Parataxis and hypotaxis are both employed, though linking elements are generally the coordinating conjunction & or the relative particle þe. It is noticeable that each verse-unit (“half-line”) tends to coincide with clauses or longer phrases. Old English element-order is also sustained. Verb-final constructions are common in subordinated
445
446
IV Middle English clauses, such as þet þeos boc rede and þa hine to monne iber. Subject-verb ordering is found in & þa scipen foren│ wide יִʒeon þare wintrede sæ, even within a clause beginning with &. Double negation is deployed in nefede he│ bern no ma; it is interesting that in this example the negator ne, with which the verb hefede has been assimilated, is immediately followed by the verb, suggesting that ne is seen as an adverb (cf. also Nefede he boten anne│ sune). Noun-phrase syntax is more archaic in this passage than in Passage 1. Inflected forms of the determiner appear, e.g. alcne│ æðele mon (alcne = ACC), for þene almiten godd (ACC), to þare sæ (DAT); phrases such as for þene almiten godd also demonstrate concord between determiner and modifying adjective. As yet the indefinite article has not emerged boten anne│ sune; includes an inflected form anne derived from the OE numeral a¯n ‘one’. In the verb phrase, the formal subjunctive is retained in þat he þeos soðfeste word יִseg│ge to sumne. The complex verb phrase has not yet fully emerged; the retention of the delayed past participle in Þa grickes hefden│ troye יִmid teone bi wonen. & þat lond│ iwest & יִþa leoden of slawen. suggests that the participle is still, as in Old English, conceived of as an adjective postmodifying the object rather than as the headword of a complex verb phrase. Example (4) is from MS London, British Library, Egerton 3309 (olim Castle Howard), the Metrical Life of St Cuthbert, dating from the middle of the 15th century; the passage is taken from the first tranche presented in MEG-C (Stenroos et al. 2011), with a few minor modifications. The language has been localized to County Durham (Laing and Lass 2007:LP13).Themostrecenteditionoftheworkdatesfromthelate19thcentury(Raine1889). (4)
[folio 18r] […] To pray for þaim we halde it waste þai haue fordone oure alde lawes And broght in newe þat na man knawes Were þai all deede it war na charge þan myght we leue all at oure large […] Þis tale to saint bede was tolde Of ane of þaim þat case beholde Þat was a trewe and symple man Þat walde noʒt lye ne feyn þan In þis forde chapitill þou sall wat gif þou rede will Before he was fourten ʒere elde he had his wittes wele in welde hende hirdmen he was sett amange he saw aungels with ioy and sange Bischop saule Aydane beere to heuen […]
28 Middle English: Syntax [folio 18v] […] And on a nyght when þai slepyd He waked in prayers as he was wont He saw with in a schort stont Come fra heuen alufsom lyght And þat with many worthy wyght þai toke a saule was clere and clene And bare it to heuen þaim betwene […] lo brethir a litil stounde J haue bene wakand on þis grounde heuen yate J . saw opyn And haly aungels lede þider in þe saule of some bischop it is þat with slyke lyght was ledd to blis (Metrical Life of St Cuthbert folio 18r–18v; Raine [ed.] 1889) ‘ “We consider it foolish to pray for them who have abandoned our old laws, and brought in new ones that no man knows; it would be no difficulty if they were all dead, then we could leave everyone unconfined […]” This tale was told to Saint Bede [sic] by one of those who observed that event, who was a true and simple man, who then wished not to lie or pretend. In this fourth chapter you shall know, if you wish to read it. Before he was fourteen years’ old he had his wits well under control; he was set among noble shepherds; he saw angels, with joy and song, bear to heaven the soul of Bishop Aidan. […] And one night when they slept he stayed awake in prayers as he was accustomed to do; he saw within a short time a beautiful light come from heaven, and that with many a noble person they took a soul who was clear and pure, and bore it to heaven between them […] “… Lo brother, I have been waking at this place for a little while; I saw heaven’s gate open and holy angels lead in thither the soul, it is, of some bishop, who was lead to joy with such light” ’. It will be immediately observed that the passage lacks punctuation as we understand it. Litterae notabiliores are deployed generally at the beginning of lines, but not elsewhere. The reason for this omission is fairly obvious: punctuation is not needed for purposes of disambiguation. In contrast to example (3), the layout of the text on the manuscript-page clarifies the verse structure: the poem is written out as verse, not like prose; and, as has often been commented upon, the choice of iambic tetrameter means that there is a persistent close correlation between verse-line and grammatical unit (see Attridge 1982: 81–82). We might note how many of the lines in example (4) begin with a conjunction (either coordinating or subordinating) or a preposition. Close examination of the passage without imposing modern punctuation practices reveals that the grammatical structure of the passage is loose, with periods linked by conjunctions in a way which modern readers would see as more appropriate for speech rather than formal written text. Many changes from the usages found in example (3) may be noted. We might examine these lines: And on a nyght when þai slepyd He waked in prayers as he was wont
447
448
IV Middle English He saw with in a schort stont Come fra heuen alufsom lyght And þat with many worthy wyght þai toke a saule was clere and clene The dominant element-order in these lines is subject-verb, even after two initial adverbials, as in And on a nyght when þai slepyd/ He waked in prayers. An exception is Come fra heuen alufsom lyght, although here rhyming constraints are probably significant. More interesting is the omission of an explicit relative pronoun between saule and was in þai toke a saule was clere and clene, a fairly common usage in northern Middle English. But, most notably, it will be observed that there is in example (4), in contrast with example (3), a distinct structure of noun and verb phrases which is much more like Present-day English. The inflectional system within the noun phrase retained in example (3) has largely disappeared; the article system is clearly in place, as in a saule. The complex verb phrase seems to be established in J haue bene wakand. The present-day configuration of the use of the subjunctive seems to be emerging in Were þai all deede it war na charge þan myght we leue all at oure large where the expression myght we seems more appropriately translated as ‘might we’ rather than ‘could we’. A formal subjunctive is retained in a conditional clause, as in gif þou rede will (cf. second person singular indicative wilt), but even here will seems to imply simple futurity and not have the implication of volition.
4 Form and function A long-standing debate amongst historical linguists has been to do with the primacy of form over function in linguistic change. Its origins lie in debates that preoccupied the neogrammarians of the late 19th century, and those who reacted against them; the discussion underpins classic accounts such as Samuels (e.g. 1972: 28–29), and remains a live issue in more recent surveys with different theoretical bases, such as (for syntax) Fischer et al. (e.g. 2000: 319). In essence, and as applied to the history of English syntax, the question is: does a change in form, such as (e.g.) inflectional loss, drive the development of (say) a more fixed element-order? Or is it that inflections are dropped because they no longer have a function, that function having been taken over by a more fixed element-order? The answer, it has been held for several years, is that both form and function are relevant for the history of syntax; and this observation may be related to Kretzschmar’s (2009) discussion of the “linguistics of speech”. Kretzschmar’s observations, though brought upto-date and of course harnessing notions deriving from his engagement with electronic corpora, can be related rather well to the famous distinction made by Samuels (1972: 139) between the “spoken chain” (Samuels’s term for Saussure’s parole) and “system” (Samuels’s term for Saussure’s langue). Samuels defined these two notions as follows: Spoken chain: the total utterances of a given group or community over a limited period, whether fully intelligible or not; System: the total of accepted and intelligible norms […] in the same group and period (1972: 139).
28 Middle English: Syntax Samuels (1972: 140) saw the relationship between spoken chain and system as iterative; variants may be selected “according to current requirements of the system for the maintenance of equilibrium and of the level of redundancy” or they “may occur in such quantity that the selection of other minority-variants is no longer in question. They are imposed on the system, and the system is thereby altered”. Such notions still resonate in current discussions of the importance of “frequency” (as for example in the essays collected in Bybee and Hopper 2001). It seems likely, therefore, that the next major advances in the study of historical syntax in general, and of ME syntax in particular, will take place through the analysis of large bodies of data, bringing quantitative as well as qualitative insights to bear. Such approaches are already being adopted by various projects. (Several projects are currently under way which approach the study of ME syntax using corpora: most notably at Helsinki, but also at York and Manchester in the UK, at Amsterdam and Nijmegen in the Netherlands, at Kyoto in Japan, at Pennsylvania and Toronto in North America, and at many major institutions in Spain, Italy and Germany. A list of relevant websites appears immediately before the List of References, Section 5.) This meshing of theoretical “linguistic” insights with older data-focused “philological” approaches seems to be a fruitful and exciting way forward for the subject, and underpins the discussion offered above.
5 References 5.1 Electronic Resources Laing, Margaret and Roger Lass. 2007. A Linguistic Atlas of Early Middle English 1150–1325 (LAEME). Edinburgh: The University of Edinburgh. See http://www.lel.ed.ac.uk/ihd/laeme1/laeme1.html McSparran, Frances (ed.). 2006. Middle English Compendium. Ann Arbor: University of Michigan Press. http://quod.lib.umich.edu/m/mec/ Nevalainen, Terttu, Irma Taavitsainen, and Sirpa Leppa¨nen. 1998–. The Research Unit for Variation, Contacts and Change in English (VARIENG). Department of English, University of Helsinki. http://www.helsinki.fi/varieng/index.html Stenroos, Merja, Martti Ma¨kinen, Simon Horobin, and Jeremy Smith. Version 2011.2 The Middle English Grammar Corpus (MEG-C). http://www.uis.no/research/culture/the_middle_english_ grammar_project/meg-c/
5.2 Print References Allen, Rosamund (trans.). 1992. Lawman: Brut. Dent: London. Attridge, Derek. 1982. The Rhythms of English Poetry. Longman: London. Benskin, Michael. 1982. The letters and in later Middle English, and some related matters. Journal of the Society of Archivists 7: 13–30. Benskin, Michael. 2011. Present indicative plural concord in Brittonic and Early English. Transactions of the Philological Society (Special Issue on Languages of Early Britain, ed. by Stephen Laker and Paul Russell) 109(2): 158–185. Brinton, Laurel J. 1988. The Development of English Aspectual Systems: Aspectualizers and Postverbal Particles. Cambridge: Cambridge University Press. Brook, G. L. and R. F. Leslie (eds.). 1963. Laʒamon: Brut. (Early English Text Society, 250, 277.) London: Oxford University Press. Bybee, Joan and Paul J. Hopper (eds.). 2001. Frequency and the Emergence of Linguistic Structure. Amsterdam/Philadelphia: John Benjamins.
449
450
IV Middle English D’Ardenne, S.R.T.O. (ed.). 1961. Þe Liflade and te Passiun of Seinte Iuliene. (Early English Text Society, 248.) London: Oxford University Press. Day, Mabel (ed.). 1952. The English Text of the Ancrene Riwle (Cotton Nero A. xiv). (Early English Text Society, 225.) London: Oxford University Press. Denison, David. 1993. English Historical Syntax. Longman: London. Dobson, E. J. (ed.). 1972. The English Text of the Ancrene Riwle (Cotton Cleopatra C.vi). (Early English Text Society, 267.) London: Oxford University Press. Fischer, Olga, Ans van Kemanade, Willem Koopman, and Willem van der Wurff. 2000. The Syntax of Early English. Cambridge: Cambridge University Press. Iyeiri, Yoko. 2001. Negative Constructions in Middle English. Fukuoka: Kyushu University Press. Jones, Charles. 1988. Grammatical Gender in English 950 to 1250. London: Croom Helm. Kretzschmar, William A. 2009. The Linguistics of Speech. Cambridge: Cambridge University Press. Laing, Margaret. 1993. Catalogue of Sources for a Linguistic Atlas of Early Medieval English. Cambridge: D. S. Brewer. McIntosh, Angus, Michael Samuels, and Michael Benskin, with Margaret Laing and Keith Williamson. 1986. A Linguistic Atlas of Late Mediaeval English (LALME). Aberdeen: Aberdeen University Press. Millett, Bella (ed.). 2005. Ancrene Wisse. (Early English Text Society, 325.) Oxford: Oxford University Press. Parkes, M. B. 1992. Pause and Effect: Punctuation in the West. London: Scolar. Raine, James (ed.). 1889. The Metrical Life of St Cuthbert. London: Surtees Society. Samuels, M. L. 1972. Linguistic Evolution. Cambridge: Cambridge University Press. Seymour, M. C. (ed.). 2002. Mandeville’s Travels: The Defective Version. (Early English Text Society, 319.) Oxford: Oxford University Press. Shepherd, Geoffrey (ed.). 1972. Ancrene Wisse: Parts Six and Seven. Manchester: Manchester University Press. Smith, Jeremy J. 1996a. An Historical Study of English. London: Routledge. Smith, Jeremy J. 1996b. Language and style in Malory. In: Elizabeth Archibald and A. S. G. Edwards (eds.), A Companion to Malory, 97–113. Woodbridge: Boydell and Brewer. Tolkien, J. R. R. 1929. Ancrene Wisse and Hali Meiðhad. Essays and Studies by Members of the English Association 14: 104–126. Tolkien, J. R. R. (ed.). 1962. The English Text of the Ancrene Riwle: Ancrene Wisse, edited from MS. (Early English Text Society, 249.) London: Oxford University Press. Warner, Anthony. 1982. Complementation in Middle English and the Methodology of Historical Syntax. University Park: Pennsylvania State University Press. Warner, Anthony. 1992. English Auxiliaries: Structure and History. Cambridge: Cambridge University Press. White, Hugh (trans.) 1993. Ancrene Wisse. Harmondworth: Penguin.
Jeremy J. Smith, Glasgow (UK)
29 Middle English: Semantics and lexicon 1 2 3 4 5
Introduction Semantics Structure of the lexicon Summary References
Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 450–466
29 Middle English: Semantics and lexicon
Abstract The lexicon of Middle English has been approached from a variety of perspectives. The semasiological approach begins with the set of lexemes and investigates what they mean and what they meant at different times during the period in question. The lexicon has also been approached from the opposite direction: the onomasiological approach takes a set of objects or concepts as the starting point and investigates what words were used to express them. This chapter considers the Middle English lexis from both these points of view, beginning with semantic changes and then examining collocational studies and investigating how we understand Middle English vocabulary within the context of the evidence provided by historical lexicography. The second part of the chapter is concerned with the structure of the Middle English lexicon, examining word formation and then briefly considering word borrowing and its effects on the structures of word fields in Middle English.
1 Introduction A number of approaches have been taken to the vocabulary of Middle English (ME). The most salient of these from the position of post-medieval readers of ME texts is the semasiological, in which investigation focuses on a given set of lexemes, asking what they mean and, more specifically, what they meant at different times during the period in question. The issues arising from this endeavor include: what changes in meaning are found in the history of a particular term, and what motivations can be traced for those changes. For answers to questions framed within the semasiological approach, we may turn to historical dictionaries such as the Oxford English Dictionary (OED) (Simpson [ed.] 2000–) and the Middle English Dictionary (MED) (Kurath et al. 1952–2001). The policies that lie behind the major works of historical lexicography will form part of my discussion about the semantics and lexicon of Middle English. There is also a considerable body of work which begins from the onomasiological standpoint. This approach starts with a set of objects or concepts and investigates which words have been used to express them. One project investigating Old and Middle English from an onomasiological point of view is the Lexis of Cloth and Clothing in Britain c.700–1450 project (Owen-Crocker et al. 2006–), the aim of which is to gather together all the vocabulary which named items of dress as well as textiles, leather, and the techniques involved in making clothes (sewing, weaving, etc.), and to use the resulting database of terminology to attempt more precise designations for manuscript illustration, grave art and archaeological remains. Another resource, the recently completed Historical Thesaurus of the Oxford English Dictionary (HTOED) (see Kay et al. 2009) guides readers approaching the lexicon from either the semasiological or the onomasiological direction, since it offers both a conceptual classification of the lexis of English, arranged hierarchically and within the categories and sub-categories chronologically, and an index. This means that researchers may consider which words were available for a particular concept at a specific moment in time, or they may begin with the word and investigate which conceptual groupings it falls into at different points in its history of usage. Example (1) from the HTOED’s data shows a small onomasiological investigation: these are the terms with first usage in the ME period for the concept of HEALING/CURE:
451
452
IV Middle English (1)
Healing/cure n: bot(e)ning 1303–c1315 · recover 1303–1631 · curation c1374–1677 · mending c1375 (Scots) · warishing c1386–c1440 · recovery 1387/8–1686 · cure 1393 (also fig) – · sanation · c1440–1697 (also fig) By contrast, (2) indicates a semasiological investigation: these are the headwords of the lexical categories, in some cases subcategories, in which the term recovery appears in the HTOED with the dates of usage in the particular sense: (2)
Healing/cure 1387/8–1686 .restoration to health 1590–1774 .a cure/remedy 1620–1761 Recovery 1606– .of one’s health 1568–1678 Habits & actions ..return/capture of ringed/tagged animal 1909– Reclamation 1853 Restoration .return to a previous better state 1932 .restoration to sound/proper/normal state 1669 Recovery from misfortune/error etc. 1525– Change of direction, reversion .of material things 1885– Posture ..act of regaining original position 1876– Reaching a point/place .arrival c.1540–1653 Respect …act of rising after recovering 1712–1867 Obtaining/acquiring .back/again 1555–1863 ..of territory, etc. 1555–1788 ..possibility of recovering something 1538 .of one’s health, etc. 1568– ..again 1771 Victory in arms ..recovery of territory 1555– Fine 1479 Claim at law .fact of succeeding in claim 1472/3 (Law) Types of transfer ..recovery 1515–1741 Reform/amendment/correction 1593–1853 Arrival c1540–1653 Propelling boat by oars/paddle/pole ..recovery of oar 1856 Construction and servicing aircraft/spacecraft .retrieval of spacecraft/satellite after flight 1949– General/industrial manufacturing processes
29 Middle English: Semantics and lexicon ..extracting 1885– Constructing/working with wood .ratio of final product to log volume 1958– Payment of debt .collecting debts 1745 Profit 1931– Indebtedness ..recovering a debt 1745; 1922 In this example we see that the focus is the lexical item, which appears to have had a long life in the language from its first appearance in Middle English through to its continuing use in Present-day English. The list of meaning categories into which the term falls at different points in its history offers a beginning place for thinking about which categories have recovery as a prototypical term, and where the headword offers an idea which is on the periphery of the variational space of the term. We can see which of the occurrences of recovery occur within the categories which appear below headwords and which are found within subcategories further down the lexical hierarchies. In the excerpt from the thesaurus classification of the semantic category HEALING/CURE, the different forms recover and recovery, as well as the date ranges for their usage, suggest that ‘healing/cure’ was for a long time a core sense for the term.
2 Semantics 2.1 Semantic change What the semasiological investigation highlights is the process of semantic change. Smith (1996) sets out three stages of semantic change which may happen in any order: the conceptual meaning of a word moves from one part of its variational space to another; one (or more) conceptual or associative meaning(s) of a word within its variational space is (or are) dropped; a word develops a new conceptual or associative meaning and thus extends its variational space. A stage in the process of semantic shift is indicated by the presence of polysemy, and we can point to examples within the ME period which indicate the parameters of the variational space of individual lexical items, such as the senses ‘claim at law’ and ‘fine’ for the term recovery. Given the possibilities of movement within variational space, including shifts in register and connotation as well as meaning, it is not surprising that semantic change is evident in the ME period. It has been suggested that metaphorization may be one of the main mechanisms of semantic change (Schendl 2001: 30). This is generally considered alongside another type of meaning extension, metonymization, though neither has been considered in detail in discussions of ME vocabulary. Metaphorization is defined as the understanding of one element in a conceptual structure in terms of an element of another conceptual structure. An example is provided by the sense history of temporal while ‘during the time that’ which develops into concessive while ‘although’ within the ME period. Metonymization is generally understood in terms of the relationship part-whole, generally within the domain of physical contiguity. An example is offered by the term keel ‘the lowest longitudinal timber of a ship or boat’ and ‘a small vessel for loading ships, a
453
454
IV Middle English lighter’ (see Traugott and Dasher 2002: 28). Although both the OED (Simpson [ed.] 2000–) and MED (Kurath et al. 1952–2001) classify these two senses of keel as different nouns, they indicate almost complete overlap in their dates of usage. This evidence indicates that metonymization is a more complex notion than the idea of a development that takes the term from denoting a part to denoting the whole might suggest. Processes of semantic change which have traditionally received more attention are narrowing and widening; for example, in Old English (OE) deor meant all kinds of wild creatures, but by the mid-14th century the term was rarely used of wild animals in general and had become restricted to the modern sense ‘deer’. This semantic development is described as ‘narrowing’ or ‘specialization’. The opposite process is traceable too: in Old English the term brid denoted a young bird, such as a chicken or eaglet. According to the OED (Simpson [ed.] 2000–), by the ME period it had extended its range of meaning so that it designated ‘the young of other animals’ (from 1388); ‘a young man, youngster, child, son’ (from a1300); ‘a maiden, girl’ (from a1300); and ‘any feathered vertebrate animal’ (from a1225). Another set of widely observed processes are those of amelioration and pejoration; for example, in Old English, prætig meant ‘cunning, crafty’ but from just before the middle of the 15th century we find it used, in particular of women or children, with the sense ‘attractive and pleasing in appearance’: the sense history of this term thus provides an example of amelioration. An interesting instance of pejoration is offered by the adjective crafty: Middle English inherited from Old English the senses ‘strong, powerful, mighty’, listed first in the OED (Simpson [ed.] 2000–) but labeled ‘Obs. rare’ (presumably not in the medieval period), and ‘skilful, dextrous, clever, ingenious’. The meaning of the term shifted, however, to ‘Of persons or their faculties, etc.: Skilful in devising and carrying out underhand or evil schemes; cunning, artful, wily’ and ‘Of actions, etc.: Showing craft or cunning’. The time lines showing first and last usage, as derived from the OED (s.v. crafty adj.), look like (3): (3)
1. Strong, powerful, mighty: c893; 1340 2. Skilful, dexterous, clever, ingenious (of persons or their faculties, etc): 971–1877 (of things, actions, etc.: showing skill or cleverness; skilfully wrought): a1000–1599 3. In bad sense (of persons or their faculties, etc.): skilful in devising and carrying out underhand or evil schemes; cunning, artful, wily: 1386–1852 (of actions, etc.): Showing craft or cunning: a1225–1855 –
The MED (s.v. craftı¯ adj.) adds a little more detail about the individual senses, but tells essentially the same story (4): (4)
1. 2. (a) (b) (c)
Strong, powerful: ?a1300 Skillful, clever, learned: c1150(OE)–a1475(?a.1430) Skillfully done or made; intelligent, learned, subtle: a1375–a1500(c.1386) according to the rules of an art or science, artful, scientific: ?c1425–c1450 (1369) 3. (a) Skilled in a trade or craft; skilled workman, artisan, mechanic, builder, designer; also fig: c1275(?a1200)–(1463–4)
29 Middle English: Semantics and lexicon (b) A member of a gild; esp., one of the master craftsmen: c1400(c1378)– a1525(?1474) (c) Showing craftsmanship, well-built, well-made, workmanlike: c1275 (?a1200)–c1600(c1350) (d) Requiring skill, difficult: c1450 4. Sly, cunning, tricky, deceitful; – of persons or actions: c1225(?c1200)–1607 (?a1425) The meaning history of this term points to the role of pragmatics in semantic change. We find pragmatic considerations in play in the area of the replacement of vocabulary where register (the sense that speakers have that a term has attached to it the prestige associated with the language from which it has been borrowed) is the motor for change. This idea is a powerful one; we shall return to it when we come to look at the structure of the lexicon, but it is also worth considering it in relation to semantic change. The various classifications of shifts in meaning, such as amelioration and pejoration, offer a sense of order in handling meaning change; it has been noted, however, that because they operate with selective and abstracted data and disregard the mechanisms of change, these classifications cannot claim a place in a history of the language (Burnley 1992: 487; see also Schendl 2001: 29–30). More recently, more overarching arguments for regularity in semantic change have been made. The changes undergone by the term crafty look as if they were motivated by speakers through a process of the pragmatics of the conventionalizing of implicatures. The implied meaning when the term was used in conversation would seem to be that someone who is clever is in a position to ‘get one over’ on you, and is someone to be wary of. Conventional implicatures arise when certain conventional meanings which have become attached to certain words in a culture do not have to be inferred from the conversational context: here, crafty shifted from having the sense of physical strength, to being intellectually able, to the use of that ability to trick others. This process seems to support arguments that there are several unifying threads in recurring patterns of semantic change. One of these threads is the tendency for meanings to undergo subjectification – they come to express grounding in the speaker or writer’s perspective – and, ultimately, intersubjectification – they come to express grounding in the relationship between speaker/writer and addressee/ reader (cf. Traugott and Dasher 2002; Traugott, Chapter 11).
2.2 Collocation-based studies Some areas of the lexicon have acted as prompts to investigations which focus on collocation as a means of discovering more about meaning, register, and the attitudes of speakers in the ME period. These are not exactly semantic field studies, which we will consider below in relation to the structure of the lexicon of Middle English and the effects of extensive borrowings in the latter part of the ME period. The authors of studies focusing on collocation are not interested in examining all the lexical items in use to express a particular concept. Rather, such investigations offer onomasiological and semasiological investigations of conceptualization and word meaning in use in Middle English, often in the works of one particular writer. One area which has come under particular scrutiny within ME lexical and semantic scholarship is that of the language of dreaming, and this topic offers a representative example of this particular approach. An
455
456
IV Middle English early study of the vocabulary of dreaming in Middle English gives a flavor of the scholarship of the time; it concludes that in ME there was a greater variety of words than in Old English, and that the ‘dream’ of dreme and dremen arrived around 1250 and supplanted earlier words during the period 1350 to 1400. A few ME terms occur rarely: reuelacioun, aperans, shewynge, oracle, metels, dremeles, fantom, and miracle. Three nouns are used frequently: sweven (< OE), vision (Romance loan), and dreme (whether from Old English or perhaps from another source or influenced by Old Norse draumr, dreymir). Verbs are: swefnian, meten and dremen. Ehrensperger (1931) argues that there appears to be no distinction between either the nouns or the verbs of dreaming in terms of shades of meaning and it remains mysterious as to exactly how and why dreme and dremen came to supplant earlier terms. More recently we find examinations of the same set of terms, some explicitly based on Ehrensperger’s (1931) work. Scholars have traced the development of specific words from Old English onward and demonstrated that fundamental changes took place in the vocabulary of dreaming during the ME period. In a common move, a great many scholars have considered Chaucer’s usages in this area. Chaucer’s vocabulary is said to have looked forward in that, in comparison with his contemporaries, he used the noun drem proportionately more than the older noun sweven but both he and Gower use the verb meten (said to be older) twice as readily as the ‘newer’ dremen. However, the reasons for lexical change still remain mysterious: all words deriving from OE swefn- and mætdisappear, and are replaced by drem- in Middle English; perhaps meten had too many homonyms and sweven was ambiguous (Fischer 1996). A literary study that attempts to capture the nuance of the choice of dream terminology comments on Chaucer’s lexical choice at one particular moment in his poem Troilus and Criseyde. Deploying philology to argue against the views of most critics of the poem that the scene of Troilus’s presumed inactivity in his bedroom represents the end of his solipsism and the beginning of his conversion to the ethos of devotion, Moore (1998: 50) suggests that Chaucer’s use of the verb meten in this scene makes the reader interpret Troilus “not as a self-absorbed daydreamer but as someone in deep, focused, creative discovery”. The most recent work on the ME dream vocabulary takes a more corpus-based approach. Łozowski (2005) begins by observing that establishing meanings of words is a basic task as well as a major challenge for ME semantics. His study attempts to carry out detailed analyses of contextual variation in order to uncover Chaucer’s use of the two dream verbs, meten and dremen, and demonstrate Chaucer’s subtle and intentional attempts to keep the two terms apart. Drawing on cognitive semantics, Łozowski argues that, as their development in both Old English and Middle English indicates, meten and dremen cannot be reduced to any common-denominator (componentially-delimited) type of denotation and presented as “cognitive synonyms”. Equally, they cannot be differentiated simply in terms of (prototype-based) radial categories and thus given as differently profiled configurations of one and same set of attributes (Łozowski 2005: 125). Łozowski (2005: 129) first examines distributional and functional differences, including such features as there being no personal mætan vs. no passive swefnian, and only accusative objects for swefnian vs. only human subjects for mætan. Close analysis of 70 contexts in Chaucer, Gower’s Confessio Amantis, and Langland’s Piers Plowman reveals no evidence for Burnley’s (1989: 36) inclusion of dremen among the verbs that “are frequently used by Chaucer in impersonal constructions”: the only two impersonal uses of dremen in Chaucer are Sir Thopas line 787 (“Me
29 Middle English: Semantics and lexicon dremed al this nyght”) and The Romaunt of the Rose line 51 (“That it was May, thus dremed me”). Łozowski also disproves the suggestion by Fischer that although Chaucer had extensive knowledge of dreams “he did not employ this knowledge to use the vocabulary at his disposal to make fine semantic distinctions” (Fischer 1996: 255; Łozowski 2005: 128). Furthermore, Chaucer was no more progressive in using dremen than Gower, nor was he prepared to use dremen in the sense ‘have a sleeping vision’. So, although “Chaucer points forward while Gower points backwards”, as suggested in Ehrensperger (1931) with regard to the dream nouns (sweven and drem), both authors use the “older” verb meten twice as readily as the “newer” dremen. Moreover, rather than the simple quantitative meten/dremen ratio, what really stands for the newer trend in the use of dream-verbs is the from-impersonal-to-personal shift and in that respect, Chaucer, with his frequent uses of meten in impersonal constructions, lags behind not only Gower, but also Langland (Łozowski 2005: 131–132). These approaches to the ME lexicon of dreams and dreaming offer a sense of the questions that linguists (and literary critics) are asking in the face of our distance from speakers of Middle English and the difficulties associated with gaining an understanding of the connotations, range of associative meanings, and register of the vocabulary for those speakers. It has been pointed out that our evidence for the range of registers which were available in the vernacular during the late medieval period is limited: we have to make do with the written materials which have survived the vagaries of time, or with interpreting the discussions of contemporary commentators. Scholars of the period do not have the advantages available to present-day dialectologists and sociolinguists; many groups in society were illiterate, and this means we have no direct access to their language (Smith 2006: 125). The problem is nicely delineated by Diller (2005: 111) in his work on Chaucer’s emotion lexicon: having offered a definition for emotion, he comments that the definition offered is “unashamedly modern”. The task is to identify those cases when a ME word refers to what would be called an emotion or emotions in general in the 21st century; the expectation is that there may be no ME term co-terminous with our modern word emotion. One of the first challenges to addressing this potential anachronism is that the meaning of ME mood is identical neither with that of Modern English (ModE) emotion nor with that of ModE mood. Emotion is, linguistically speaking, somewhat like illness: we do not say that somebody’s illness changed (e.g. from whooping-cough to pneumonia) any more than we would say that their emotion changed (e.g. from wrath to mercy). On the other hand, most speakers of Modern English seem to find it acceptable to say that somebody’s mood changed, just as their temperature or their complexion might change. There also seems to be a “criterial” or “expected” feature “not caused by an antecedent” which for many speakers is present in ModE mood but was absent in ME mood and which the OED fails to note (Diller 2005: 113). Quotations from Chaucer’s Boece and Troilus and Criseyde indicate that there are bound to be times when the pigeon-holing of a word-token into a lexical sense becomes arbitrary. The lexicographer has to live with such arbitrary decisions. The historical linguist, who wants to reconstruct the intuitions of native speakers of bygone ages, should be more interested in the relationship between the senses and the principles that lead from one sense to the next (Diller 2005: 117). As we shall see in the next section, however, historical lexicographers struggle with this issue just as historical linguists do.
457
458
IV Middle English
2.3 Historical lexicography and lexicology This chapter is not primarily concerned with Middle English as a literary language; nevertheless, as the word studies discussed above make clear, understanding the lexis of Middle English generally involves readers, texts, and historical dictionaries. The MED’s early editors were aware of the two possibilities for their historical dictionary: that it should serve as a repository of information about the language as language, or function as a glossary of printed texts in Middle English (Adams 2002: 106). Kurath, who led the dictionary project from 1946 until 1961 and oversaw the first publication, appears to come down on the side of the MED’s being a storehouse of information about Middle English as a language variety, rejecting the possibility of its functioning as an inventory of glosses for lexical items in texts as inappropriate to a large-scale dictionary. He also suggests that including glosses of particular usages would cause the dictionary to give inappropriate weight to individual construals of lexical meaning, which he distinguishes from distinct meanings. Lewis (the final editor of the completed MED) is clear that the guiding editorial principle of the dictionary in the latter years of the project was to try to “capture the generality” of the language variety, but adds that the intention was that the dictionary should, at the same time, give the reader as much help as possible with difficult quotations and with subtleties of meaning (Lewis 2002: 81). Applying the information found in the historical dictionaries to the lexical items as used within texts, however, is not always straightforward. An examination of the verbs in the clauses which make up the bedroom scenes in Sir Gawain and the Green Knight offers some indications of the dilemmas (see Sylvester 2010). In the final bedroom scene, there is only one clause which seems to encode a material process intention (that is, an action undertaken deliberately) in which Gawain is the actor: his refusal of a ring that Lady Bertilak attempts to give him. The clause in question runs: Bot þe renk hit renayed (line 1821). It seems possible that Gawain accomplishes the rejection of the proffered ring by uttering a formula of refusal; that is, he might have done it by means of a verbal rather than a material process. Turning to the MED, we find that the verb reneien is defined as having two senses: the first has to do with forsaking, renouncing, or recanting religious beliefs (orthodox or heretical), or refusing to acknowledge one’s king or master. The second, for which this moment in Sir Gawain and the Green Knight provides one of the citations, is 2(a) to refuse (a gift); refuse an invitation from (sb.). Included within the definition are also (b) to forsake (an activity, a state); abandon (a place), refuse a mission to; (c) to retract (a pledge). There is also a figurative sense: (d) to withdraw (one’s heart) from (devotion to sb.). With the exception of the first two senses under (b), all of these activities seem most likely to be accomplished by words or even, in the case of (d), by thoughts. Perhaps we can assume that Gawain makes a gesture of refusal on the grounds that Lady Bertilak is represented as handing the ring to Gawain (4): (4)
Ho raȝt hym a riche rynk of red golde werkez, Wyth a starande ston stondande alofte Þat bere blusschande bemez as þe bryȝt sunne; Wyt ȝe wel, hit watz worth wele ful hoge Bot þe renk hit renayed, (lines 1817–1821)
29 Middle English: Semantics and lexicon ‘She offered him a precious ring worked of red gold with a sparkling stone set above that shone with beams like the bright sun. Know well, it was worth a great deal. But the man refused it’. A recent translation of the poem has ‘She offers him a ring’ but the idea that Lady Bertilak actually gives the ring to Gawain rather than just offering it to him is supported by the MED’s citation of the relevant line from Sir Gawain and the Green Knight in illustration of sense 4(a) of the verb rechen: To hand (sth. to sb.), give; grant (a kingdom or dwelling to sb.); also, give (sb. a kiss). Here the evidence for the meaning of a lexical item in one part of the dictionary seems to clash with the sense of another lexical item in a context in which they co-occur. This is not to say that the definitions proposed in the MED are incorrect in any way. Its editors are aware that some quotations are ambiguous and, like the OED, the MED uses a caveat like “(sometimes) difficult to distinguish from sense such and such (and vice versa)”, with variations such as “some quots. difficult to distinguish from those in sense such and such” or “difficult to distinguish from sense such and such to which some quots. may belong”; others include “may belong to”, “could be construed as”, “the precise gloss is highly contextual” (see Lewis 2004: 150). Familiarity with Middle English creates sensitivity to the semantic associations suggested by recurring collocations and colligations: as Hoey (2005) shows, lexical items may be primed to co-occur with other lexical items, and they may also be primed to occur in, or with, particular grammatical functions. Finally, however, as the editors of the MED note, Middle English often presents nuances of meaning which are almost impossible to capture in a formal definition but which a reader can perceive in the quotations. Such nuances are often vaguely sensed rather than perceived. They may seem so much like mere subjective impressions that an editor would hardly venture to include them in a definition, yet they can be very important (Kuhn 1982 [1976]: 36).
3 Structure of the lexicon The lexicon of Middle English differs considerably from that of Old English. The answers to the questions of what the lexicon of Middle English is made up of, and how it came to differ from that of its predecessor, involve the two main strategies for the introduction of new words: word formation (the creation of new words out of the existing resources of a language) and word borrowing (the introduction, in the form of different types of loans and calques, of terms originating in other languages). Of course, blending of these two strategies can occur; for instance, some scholars are engaged in the investigation of how far affixes which entered the language via borrowings from French, for example, were integrated and became productive in Middle English (see, as a case in point, Lloyd 2005 on the nominal suffixes -ment, -ance/-ence and -ation).
3.1 Word formation Over the past century ME word formation has received relatively little scholarly attention (Ciszek 2008). The two methods of word formation which are of greatest importance in Middle English are compounding and derivation. Many compounds are formed from pre-existing syntactical groups, and so at a certain point in their
459
460
IV Middle English development, fluctuation may occur between their interpretation as compounds or syntactic groups. The boundary between compounds and derivations may also be unclear; for example, the suffix -ly is related to the OE noun lic, meaning ‘form, shape, body’ and in earlier Germanic it had been a free morpheme frequently used to create noun compounds.
3.1.1 Compounding Despite the fact that compounding was less fertile than in the OE period, many of the OE types of compounding continued to be productive. Noun compounds were numerically the commonest in Old English, and many of these types continued to be productive. Those of the N + N type were especially frequent and the ME period sees the formation of a number of examples which survive up to the present day, including bagpipe (OED [Simpson, ed. 2000–] records this as first used by Chaucer in his description of the Miller in the General Prologue to the Canterbury Tales, though the MED [Kurath et al. 1952–2001] has a predating; the term certainly first appeared in the ME period), bedchamber, birthday, bloodhound, schoolmaster, and swordfish. Nouns compounded from ADJ + N in the ME period are blackberry, blackboard, grandfather, highroad, highway, and shortbread. Some new types of compound noun emerged during the ME period, most notably those in which a verbal stem is completed by a nominal which, in the underlying sentence would have acted as the subject of the verb; for example, leap-year (for fixed festivals, the year ‘leaps’ a day). In the 14th century, a type where the second element was an agent noun and the first the object of the underlying verb, which had been present in Old English but had died out, became newly productive; for example, moneymaker (1297), housebreaker (1340), soothsayer (1340), lawmaker (1380). Exocentric compounds (in which the denotation of the compound noun is not a subset of either the determiner or the head, but is included within the sense of a more general conception) developed considerably in Middle English. An example is provided by burnwater ‘smith’. Copulative compounds (in which it is not clear which element is the grammatical head since both elements refer equally to the referent) are represented in Middle English by a few 13th-century noun formations; for example, kayserr-kinng (Ormulum), stane-roche (Vices and Virtues). Compound adjectives include a type in which the first of two adjectival elements modifies a second, making fine distinctions in sense impressions, e.g. icy-cold, red-hot (1375), lukewarm (1398), light-green (1420). Combinations of a noun and past participle became productive again in the 14th century; for example, moss-grown (1300), woe-begone (1470), moth-eaten (137), book-learned (1420). Adjectival compounds formed with the past participle as head also include a type in which the determiner is an adjective or an adverb. Most extant examples date from the 14th century, such as new-born (1300), high-born (1300), and new-sown (1375). In Middle English all these types continued, but they began to be redistributed into inseparable PARTICLE + V compounds (understand, overtake); phrasal verbs consisting of VERBAL BASE + PARTICLE (take up, write up); and derived nominal compounds of the two types (outcry, write-off ). Among many examples of co-existing compound and particled verbs are flee out (1300) and outflee (1325) ‘expel, banish’; hente out (c.1400) and outhente (1450) ‘grasp, seize’, etc. The compound forþferan, which was an OE euphemism for ‘to die’, continues with this sense until the end of the 14th
29 Middle English: Semantics and lexicon century, but then develops a new sense ‘to set out’, presumably re-adopting its original sense from the particled verb fare forth, recorded from 1225 onwards. This emphasis on the particled verb as the focus of derivation is symptomatic of the change which took place during the 15th century by which formation of verbs became concentrated on the production of particled verbs, and compound verbs were no longer productive as a type of word formation. The derivation of agentive nouns from particled verbs, such as Chaucer’s reference to Troilus as ‘holder up of Troye’ also belongs to the turn of the 14th and 15th centuries (see Marchand 1969; Burnley 1992; Kastovsky 2006).
3.1.2 Derivation It has been suggested that Middle English is the starting point of a development which resulted in a restructuring of the English word-formation system through borrowing from French and Latin. This led to a system with two derivational strata, a native and a foreign one, although these partly overlap. The former is word-based and baseinvariant, whereas the latter is at least partly stem-based and exhibits morphophonemic alternations of the base; for example, with -able we find word-based allow-able, understand-able and also stem-based charit-able, navig-able, etc. The alternative forms in -ate and -acy, as in pirate ~ piracy, obstinate ~ obstinacy, and in -ant ~ -ancy, as in infant ~ infancy were adopted in Middle English from French. The origin of these non-native patterns was the borrowing of individual lexical items which were derivationally related in the source languages; for example, allow ~ disallow, arm ~ disarm, chain ~ enchain. Once a number of such pairs had been borrowed, a derivational relationship could also be established in English which could then be extended to new formations which may not have had parallels in the source language. This process is likely to have begun with individual analogical formations, leading to a pattern which was productive on a larger scale (see Kastovsky 2006). Productivity, however, is still a controversial concept in word-formation in general and in historical linguistics in particular (Ciszek 2008). It has been pointed out that for languages which are accessible only from texts, the establishment of synchronic productivity presents a problem because the main criteria for the establishment of productivity – introspection, elicitation, and acceptability judgements of neologisms – are not available. Such languages may be considered in terms of “analyzability”. This term refers to a situation in which paradigms of related forms (for example verb/noun/adjective) appear, and it becomes possible for users of the language to distinguish base from affix (Kastovsky 1985: 228–229). Non-native, especially French, patterns were assumed to have become productive fairly early. It has been argued, however, that the Romance suffixes did not become productive in the ME period at all (Dalton-Puffer 1996). Other scholars suggest late Middle English as the starting point for such productivity, especially for prefixation, when apparently a critical mass of borrowings and analogical formations had accumulated to get the derivational processes going (see Burnley 1992: 447–450). In view of the number of prefixes and suffixes entering the English language in Middle English according to Marchand (1969), the increase in productivity must have been gradual and certainly varied from affix to affix (Kastovsky 2006). In the ME period many OE prefixes had become unproductive or disappeared, paving the way for borrowing from French and Latin. The reversative sense of the prefix un-, which had been connected with ond- and on- in Old English, was developed;
461
462
IV Middle English the MED (Kurath et al. 1952–2001) cites unclose ‘to open’ used by the Gawain poet and by Gower in the sense ‘open’ or ‘unlock’ and by Chaucer in the sense ‘of a flower: to spread out, bloom’. A number of prefixes appear in Middle English, often filling semantic gaps in the English derivational system; for example arch-, co-, dis- (though this remained stylistically differentiated from the rest of the lexis until Lydgate’s use of the Latin form distrust (1430), en-, in-, mis-, non-, re- and vice- (see Kastovsky 2006: 251), where examples of usage are offered). Middle English was better supplied with derivational suffixes. Of the forty or so which existed in Old English, about three-quarters persisted into Middle English, and they were joined by numerous additions from foreign sources. The suffix -ful, which was originally used to form adjectives from abstract nouns, now also formed adjectives from verbal bases; e.g., forgetful (1382), weariful (1454). Suffixes from all the major sources of foreign influence achieved a limited productivity in Middle English. French and Franco-Latin are the most prolific sources of foreign derivational suffixes; e.g., -age and -ard. The latter entered the language via loans like buzzard and bastard and became productive as a pejorative suffix with English bases by the 13th century as evidenced by shreward (1297), dotard (1386) (see Burnley 1992; Kastovsky 2006 offers more detail and examples). Foreign sources which were of great importance to word formation in Middle English played an equally important role in phrase creation. French, in particular, contributed a large number of phrasal idioms of which verbal phrases especially have proved productive. The structure of such phrases usually consisted of a verbal operator followed by an abstract noun or adverbial phrase; thus: do homage, do mischief, do justice, make complaint, make moan, etc. They are especially common from the second half of the 14th century. A parallel tendency exists in verbal phrases based on the verbal operator get: phrases such as get grace, get mercy and get leave are recorded from 1300 onwards. The major contribution of this verb is, however, in a series of expressions with locatives: many of the type get away from, get up and get out occur in the ME period (see Burnley 1992).
3.2 Lexical borrowing and word field studies One difficulty with attempting to analyze the semantics of derivations is that almost any ME noun on an Old French base could have been adopted holistically, even if its date of adoption is later than that of the related verb. Arguments about the very concept of borrowing as applied to the language situation in medieval England suggest that even words obviously unanalyzable in English might have been analyzable to bilingual users acquainted with their simplex forms in French, which would facilitate conditions for learning new words (see Lloyd 2005). Although language contact in the ME period is properly the subject of another chapter, discussions of the formation of the ME lexicon makes it clear that it is not possible to speak of the lexis of Middle English without touching on the subject of word borrowing. It has been observed that the lexical stock of languages may contain a considerable proportion of words borrowed from one or more other languages and the historical record can help us infer which words were borrowed, from what language, and approximately when, but there is no unequivocal way of deciding when a lexical item from one language that is used during discourse in another language – whether by a single
29 Middle English: Semantics and lexicon speaker, or repeatedly in a community – should be considered a loanword (Poplack and Sankoff 1984: 99). The case of French in England following the Norman Conquest provides a clear example of the difficulties involved in distinguishing between lexical borrowing, interference, convergence, shift, relexification, etc. Many scholars now accept the view that Anglo-Norman, at first a minority language in England, became a language of education and law in widespread use throughout the 13th and 14th centuries, casting doubt on the status of many French words as borrowings from continental French in Middle English. It has been argued that both French and Latin had been developing in an English society on English soil for hundreds of years, independently of either Rome or Paris in the general field of imaginative literature and also in a wide spectrum of administrative and technical texts (Rothwell 2000: 214). We may also wish to distinguish the type of borrowing which simply fills a gap from that which becomes productive in the borrowing language. This issue goes to the heart of what has been described as a central problem in the history of both lexis and grammar; that is, the question of whether it is the “availability” (for mechanical, extralinguistic, or extrasystemic reasons) of new forms that causes semantic shifts, by differentiation from them of older forms, or whether it is the prior shift of the old form to a new meaning (by extension and limitation) which creates the “need” for a new form (Samuels 1972: 67). Thus, discussion of word borrowing leads directly to questions about the structure of the lexicon, an issue which is particularly pertinent to Middle English, the vocabulary of which is thought to be almost double that of Old English, between a quarter and a third of terms being loanwords. In his study of the lexical fields Boy/Girl-Servant-Child in Middle English, Diensberg (1985) explores the restructuring of these lexical fields in the ME period, pointing out that native words were largely replaced by Romance loans; native words which did not die out often shifted to the periphery and acquired pejorative meaning. A further example of perceived gaps in the language in the ME period is provided by McIntosh (1996) who notes that during the 15th century, heads of established families in English market towns and villages expressed growing concern with certain forms of social misconduct and found the habits of thoughts and expression with which they had grown up inadequate to identify, describe, and attack perceived new threats to their village or town. McIntosh examines the records of 178 courts that enjoyed public jurisdiction, spanning nearly every county for which good rolls survive for the period 1370 to 1599. She considers groups of offences and notes the interplay between jurors’ attitudes, words, and actions and those of the surrounding culture, observing that several of the most common Latin descriptors of scolding, including objurgator/-trix and garrulator/-trix, do not specify the type of spoken offence. Arguably, it is impossible to make valid generalizations about how new words enter the lexicon unless whole semantic fields are taken into account rather than individual lexemes only. We need diachronic examinations of the lexis associated with particular conceptual constructions in order to address the questions of how and why lexical change begins and continues. An example is provided by Fischer (1989) in his diachronic examination of the lexis which is associated with the seasons of the year and the lexis denoting the action of taking as husband and wife. He makes use of a lexical diffusion model to describe the lexical changes on the way from the prevailing use of wed to the prevailing use of marry, and from ae(w) ‘law’ to lagu ‘law’ and explains the loss of harvest ‘autumn’ and ae(w) as following on after the introduction of French loanwords.
463
464
IV Middle English This kind of study, and a host of investigations of semantic fields, such as those examining the color terminology in the ME period (for example Burnley’s [1976] study “Middle English Colour Terminology and Lexical Structure” and Biggam’s “Aspects of Chaucer’s Adjectives of Hue” [1993]) suggest that the time when lexical studies were criticized for dealing only with individual lexemes and being atomistic is at an end.
4 Summary Our understanding of Middle English comes from many sources and is not simply a matter of reading the texts that survive from the period. The proper interpretation of the semantics of Middle English comes largely from the work of historical lexicographers. We almost certainly derive our sense of what is and is not part of the lexicon of Middle English from the editors of the MED (Kurath et al. 1952–2001), who observe that Old English gradually turned into Middle English during the late 11th and early 12th centuries, and no matter where we draw the line between these two stages of English, we will be involved in contradictory situations. Some of our understanding is drawn from external histories which point out that Middle English consists of four different linguistic strata: monolingual native speakers of English (usually members of the lower social ranks), speaking local dialects with no supra-regional standard; a geographically but also socially distinct group which lived in the former Danelaw area, where we still have to reckon with partial Scandinavian-English bilingualism in the 11th and early 12th centuries, ending with the loss of Scandinavian as a means of communication but massive borrowing of basic vocabulary; those who mainly or exclusively used French in oral communication, usually members of the nobility (with the eventual learning of English, probably resulting in considerable bilingualism); speakers and writers of Latin, which had remained in use as the language of the Church and of scholarship (where it was also used as a spoken medium) and of public records. It is this lexicon, drawn from so many sources, which proved such a flexible instrument for Gower, Langland, the Gawain poet, Chaucer and all the other creative writers and administrators of the ME period whose texts are left to us.
5 References Adams, Michael. 2002. Phantom dictionaries: The Middle English Dictionary before Kurath. Dictionaries 23: 95–114. Biggam, C. P. 1993. Aspects of Chaucer’s adjectives of hue. Chaucer Review 28: 41–53. Burnley, David. 1976. Middle English colour terminology and lexical structure. Linguistische Berichte 41: 39–49. Burnley, David. 1989. The Language of Chaucer. Houndmills: Macmillan. Burnley, David. 1992. Lexis and semantics. In: Norman Blake (ed.), The Cambridge History of the English Language. Vol. II. 1066–1476, 409–499. Cambridge: Cambridge University Press. Ciszek, Ewa. 2008. Word Derivation in Early Middle English. Frankfurt: Peter Lang. Dalton-Puffer, Christiane. 1996. The French Influence on Middle English Morphology. A CorpusBased Study of Derivation. Berlin/New York: Mouton de Gruyter. Diensberg, Bernard. 1985. The lexical fields boy/girl-servant-child in Middle English. Neuphilologische Mitteilungen 86: 328–336. Diller, Hans-Ju¨rgen. 2005. Chaucer’s emotion lexicon: Passious and affeccioun. In: Ritt and Schendl (eds.), 110–124.
29 Middle English: Semantics and lexicon Ehrensperger, E. C. 1931. Dream words in Old and Middle English. Papers of the Modern Language Association 46: 80–89. Fischer, Andreas. 1989. Aspects of historical lexicology. In: Udo Fries and Martin Heusser (eds.), Meaning and Beyond: Ernst Leisi zum 70. Geburtstag, 71–91. Tu¨bingen: Gunter Narr. Fischer, Andreas. 1996. Dream theory and dream lexis in the Middle Ages. In: Ju¨rgen Klein and Dirk Vanderbeke (eds.), Anglistentag 1995 Greifswald: Proceedings of the Conference of the German Association of University Teachers of English 17, 245–257. Tu¨bingen: Max Niemeyer. Hoey, Michael. 2005. Lexical Priming: A New Theory of Words and Language. Oxford: Oxford University Press. Kastovsky, Dieter. 1985. Deverbal nouns in Old and Modern English: From stem-formation to word-formation. In: Jacek Fisiak (ed.), Historical Semantics Historical Word-Formation, 221– 261. Berlin/New York: Mouton de Gruyter. Kastovsky, Dieter. 2006. Vocabulary. In: Richard Hogg and David Denison (eds.), A History of the English Language, 199–270. Cambridge: Cambridge University Press. Kay, Christian, Jane Roberts, Michael Samuels, and Irene Wotherspoon (eds.). 2009. Historical Thesaurus of the Oxford English Dictionary With Additional Material from A Thesaurus of Old English. Oxford: Oxford University Press. Kuhn, Sherman M. 1982 [1976] On the making of the Middle English Dictionary. Dictionaries 4: 14–41. Kurath, Hans, Sherman M. Kuhn, John Reidy, and Robert E. Lewis. 1952–2001. Middle English Dictionary. Ann Arbor: University of Michigan Press. http://quod.lib.umich.edu/m/med/ Lewis, Robert E. 2002. The Middle English Dictionary at 71. Dictionaries 23: 76–94. Lewis, Robert E. 2004. Aspects of polysemy in the Middle English Dictionary. In: Julie Coleman and Anne McDermott (eds.), Historical Dictionaries and Historical Dictionary Research: Papers from the International Conference on Historical Lexicography and Lexicology, at the University of Leicester, 2002, 149–156. Tu¨bingen: Max Niemeyer. Lloyd, Cynthia. 2005. Experience or experiment? Some distinctions between French nominal suffixes in Middle English. In: Ritt and Schendl (eds.), 179–195. Łozowski, Przemysław. 2005. Polysemy in context: meten and dremen in Chaucer. In: Ritt and Schendl (eds.), 125–143. Marchand, Hans. 1969. The Categories and Types of Present-Day English Word-Formation. A Synchronic-Diachronic Approach. 2nd edn. Munich: Beck’sche Verlagsbuchhandlung. McIntosh, Marjorie K. 1996. Finding language for misconduct: Jurors in fifteenth-century local courts. In: Barbara A. Hanawalt and David Wallace (eds.), Bodies and Disciplines: Intersections of Literature and History in Fifteenth-Century England, 63–86. Minneapolis/London: University of Minnesota Press. Moore, Marilyn Reppa. 1998. Who’s solipsistic now? The character of Chaucer’s Troilus. Chaucer Review 33: 43–59. Owen-Crocker, Gale R., Louise Sylvester et al. 2006–. Lexis of Cloth and Clothing in Britain c.700–1450. University of Manchester. http://lexisproject.arts.manchester.ac.uk/ Poplack, Shana and David Sankoff. 1984. Borrowing: The synchrony of integration. Linguistics 22: 99–135. Ritt, Nikolaus and Herbert Schendl (eds.). 2005. Rethinking Middle English: Linguistic and Literary Approaches. Frankfurt: Peter Lang. Rothwell, William. 2000. Aspects of lexical and morphosyntactical mixing in the languages of medieval England. In: D. A. Trotter (ed.), Multilingualism in Later Medieval Britain, 213– 232. Cambridge: D. S. Brewer. Samuels, M.L. 1972. Linguistic Evolution with Special Reference to English. Cambridge: Cambridge University Press. Schendl, Herbert. 2001. Historical Linguistics. Oxford: Oxford University Press. Simpson, John (ed.). 2000–. The Oxford English Dictionary. 3rd edn. online. Oxford University Press. www.oed.com/
465
466
IV Middle English Smith, Jeremy J. 1996. An Historical Study of English: Function, Form and Change. London: Routledge. Smith, Jeremy J. 2006. From Middle to Early Modern English. In: Lynda Mugglestone (ed.), The Oxford History of English, 120–146. Oxford: Oxford University Press. Sylvester, Louise. 2010. The roles of reader construal and lexicographic authority in the interpretation of Middle English texts. In: Margaret Winters, Heli Tissari, and Kathryn Allen (eds.), Contributions to Historical Cognitive Linguistics: Syntax and Semantics, 197–222. Berlin/New York: Mouton de Gruyter. Tolkien, J. R. and E. V. Gordon (eds.). 1967. Sir Gawain and the Green Knight. Rev. by N. Davis. Oxford: Clarendon Press. Traugott, Elizabeth and Richard Dasher. 2002. Regularity in Semantic Change. Cambridge: Cambridge University Press.
Louise Sylvester, London (UK)
30 Middle English: Pragmatics and discourse 1 2 3 4 5 6 7 8 9 10 11
Introduction Information structure and word order change Auxiliaries Degree modifiers and focus particles Pragmatic markers Speech acts Politeness strategies Pragmatic and discourse properties of genres and text-types Register Middle English as a period of transition References
Abstract Among changes in the Middle English period that involve pragmatics, two stand out in particular because of their significant effects on the later history of English. One was the shift from information-structure-oriented word order in Old English to “syntacticized” order in Middle English; this in turn led, at the end of the period, to new strategies for marking topic and focus in special ways. The second change was the development of auxiliary verbs in contexts where implied abstract temporal, modal, or aspectual meanings of certain concrete verbs became salient. The period is also characterized by the appearance of a large set of new discourse types, from romances to drama, scientific writings, and letters. These give a partial window into differences between oral and literate styles at the time.
1 Introduction Work on pragmatic and discourse factors in Middle English differs from such work on Old English for two main reasons: internal linguistic changes, and the emergence of new discourse types or genres that require methodologies different from those in Old Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 466–480
30 Middle English: Pragmatics and discourse English. Linguistic changes that have received particular attention from the perspective of pragmatics are changes in word order, and the development of auxiliaries. Most important among the new genres in the early period were romances and chronicles like Sir Orfeo, King Horn, and Layamon’s Brut. Most were based on French models, and although these narratives had some similarities with OE epics, they had different conventions. Later in the period there were drama, personal and business letters, vernacularized scientific works, and several other text types. Issues of pragmatic markers, speech acts, and the orality-literacy continuum dominate much of the research. The data bases used have been primarily electronic corpora. These include most notably the ME part of the Helsinki Corpus of English Texts (HC) (Rissanen et al. 1991), and the Corpus of Early English Correspondence (CEEC) (Nevalainen et al. 1998). Individual texts that have received particular attention from the perspective of pragmatics and discourse factors include Chaucer’s works and the Paston Letters.
2 Information structure and word order change It is generally accepted that word order in Old English was predominantly verb-final with verb-second in main clauses, while word order from early Middle English on is largely verb-medial, and syntactic verb-second was mostly lost by the end of the period. Still debated is the question why and how these changes came about. Arguments have been put forward highlighting structural factors such as loss of case morphology, changes in the status of personal pronouns, prosody changes, and parameter shifts, or sociolinguistic factors, such as the influence of Scandinavian (for some views see Taylor and van der Wurff 2005; Fischer and van der Wurff 2006). The word order shift to verb-medial has been interpreted as a change from syntax governed largely by information structure factors such as topic, focus, contrastive topic, referent accessibility, animacy, etc. (see Molenki, Chapter 19; Lenker, Chapter 21), to syntax mainly governed by syntactic relations such as subject – predicate (e.g. Kemenade and Los 2006; Kemenade 2009; Kemenade and Los, Volume 2, Chapter 94). Kemenade has proposed that this kind of change also affected subclauses, as evidenced by word order patterns associated with adverbial markers such as þa and þonne ‘then … when’ and the position of personal pronouns. She regards the steep decline in early Middle English of þa and þonne in subclauses as evidence of a reorganization of relatively paratactic into hypotactic clause structure. Questions that arise with respect to Kemenade’s analysis include to what extent þa and þonne in subclauses should be understood to exemplify a) marking of the contrastive topic and of the boundary of “given” information, or b) general inferential meaning such as is found in Present-day English (Schiffrin 1992). Another is how the decline of these markers ties in with the text-organizing function of when in its early stages (see Sections 5 and 9 below). First constituents in verb-second languages are not restricted, so “any constituent can appear as unmarked, old information” (Los 2005: 274). The decline of syntactic verb-second in main clauses is followed in later periods by the restriction of syntactic subjects to unmarked topics (in the sense of what the clause is about, and of givenness). Los hypothesizes that with the loss of verb-second, old information
467
468
IV Middle English could be expressed in English only by subjects, and the development of passive to-infinitival “exceptional case marking” was developed in late Middle English as one strategy to expand ways to create subjects (Los 2005: 274). An example, as in (1), is: (1)
and ye can, axe the probate of my fadyrs wyll to be geuyn yow (Paston Letters [Davis 1971] 338, ll. 41–42; Los 2005: 246) ‘if you can, ask-for the probate of my father’s will to be given to you’
Building on work by Los, Seoane (2006) hypothesizes that “long passives” (passives with by-phrases) contributed to the rise of this new restriction on subjects. While her focus is on Early Modern English, she argues that in Late Middle English and Early Modern English triggers for the reorganization of the first constituent of the clause are pragmatic factors such as given-new and the degree of definiteness of the subject and the by-phrase, together with syntactic factors such as end-weight. Semantic factors that are significant in Present-day English, such as animacy and human features, played no role in Late Middle English.
3 Auxiliaries What have been called “pragmatic” approaches to modality have principally drawn on two factors: the speech act functions of modals, and implicatures or invited inferences. Gotti (2005) investigates the use of shall and will in Middle English from the perspective of their speech act function, specifically whether they express prediction, prophecy, or assurance. He regards these as pragmatic functions since they pertain to contextual elements of the speech event. One of the most detailed studies of the emergence of auxiliaries in Middle English from the perspective of implicatures is Danchev and Kyto¨’s (1994) analysis of the rise of be going to in the very late ME period and its conventionalization as an auxiliary by about 1700. Another is Eckardt (2006: Chapter 4), which builds on their work. Danchev and Kyto¨ point to the multiple interpretations that many examples can be given. These are what would now be considered “bridging” examples, i.e. examples in which at least two meanings are possible, but one is “only contextually implicated” (Evans and Wilkins 2000: 549): in this case intentional future is implicated, motion is semantically coded. The famous example from the Monk of Evesham in (2) has long been thought to be the first where intentional future is clearly implicated by the motion-with-a-purpose construction: (2)
Therefore while thys onhappy sowle by the vyctoryse pompyse of her enmyes was goying to be broughte into helle for the synne and onleful lustys of her body (1482 Revelation to the Monk of Evesham; Danchev and Kyto¨ 1994: 61) ‘Therefore, while this unhappy soul by the victorious spectacles of her enemies was going to brought into hell for the sin and unlawful lusts of her body’
The Monk’s revelation was translated from Latin. To this example may be added another, slightly earlier, translation, this time from Arabic, (3):
30 Middle English: Pragmatics and discourse (3)
Also ther passed a theef byfore alexandre that was goyng to be hanged whiche saide O worthy king saue my lyf for I repente me sore of my mysdedes. (1477 Mubashshir ibn Fatik, Abu al-Wafa’ [11th century], Dictes or sayengis of the philosophhres) ‘Also there passed a thief before Alexander that was going to be hanged who said “O worthy king, save my life, for I repent me very much for my misdeeds” ’
Like (2), this example is passive, hence the agentivity associated with go as a motion verb and the purposiveness associated with to are demoted. Nevertheless, both examples suggest that motion was intended: in (2) the soul is, according to medieval beliefs, roaming the streets, and in (3) the thief is said to have been ‘passing’ in front of Alexander. So both subjects (soul and thief) were en route, though not intending to be brought into hell, or hanged; the intention rested with their captors. Krug discusses the appearance in Middle English of want ‘lack’, borrowed from Old Norse with impersonal syntax. He too shows “context-induced inferencing” (or “bridging”) from ‘lack’ to ‘desire’ in Middle English, but emphasizes that the desiderative, pre-modal meaning was not semanticized until the 18th and 19th centuries (Krug 2002: 142–143). A hitherto largely unstudied modal verb construction is the topic of Kyto¨ and Romaine (2005). Be/have like to + V ‘imminently likely to V’ began to be used in Late Modern English as an “avertive” modal auxiliary marking “action narrowly averted”. Kyto¨ and Romaine show that the new meaning developed in syntactic-semantic contexts including past tense, conditional if or but, and infinitive verbs with semantically negative “prosody” (preferred semantic collocation, see Stubbs 1995). They also discuss the role of invited inferencing, specifically the implicatures of counterfactuality and narrowly averted eventhood, in these contexts, as illustrated by (4): (4)
if he had abedyn at hom he had be lyke to have be fechid owte of his owyn hows, for the peple þer-abowgth is sore mevod with hym. (1453 Paston Letters; Kyto¨ and Romaine 2005: 3) ‘if he had stayed at home he would have come close to being fetched out of his own house because the people in those parts are very upset with him’
4 Degree modifiers and focus particles The historical study of degree modifiers with reference to their pragmatics goes back at least to Stoffel (1901). Degree modifiers include “intensives” (also known as “boosters”) (e.g. very, purely, even) and “downtoners” (e.g. rather, pretty, quite). They answer ‘how much, to what extent?’ and scale their heads. This means they typically collocate with gradable, or unbounded adjectives (Paradis 1997). They derive from lexical items with less scalar, often more concrete, meanings (very < ‘tru(ly)’, even < ‘exact(ly)’, ultimately < ‘even(ly), smooth(ly)’, and rather < ‘sooner’) (Traugott 2006). In several cases (e.g. very, even) they derive from “particularizers” or “restrictors” that can only occur with bounded heads (e.g. exactly, completely in Present-day English) (Nevalainen 1994; Paradis 2000). Many boosters developed in the EModE period, especially those derived from qualitative adverbs, such as terribly (Peters 1994), but some developed in Middle English. In this period their status as adjectives or adverbs is often unclear. There
469
470
IV Middle English was also significant reorganization within the set of available boosters. Me´ndez-Naya (2007), for example, discusses the development in Middle English of right as an intensifier. While it was favored in the later ME period over two intensifiers that were inherited from Old English: wel ‘well’ and swiþe ‘very’, it competed with a third, ful. Right won out in Early Modern English but then declined (193) and became specialized to an honorific (e.g. right honorable, þe right excellent high and mighty prince) (204). Me´ndez-Naya notes that in ME right has primarily positive semantic prosody and occurs with bounded totality markers like completely, and suggests it originates in ‘exactly, precisely’ (205). The term “intensifier” has been used not only to refer to degree adverbs that scale their heads up but also for “focus particles” which activate a scale for their heads. These heads are typically nominal, not adjectival, as in Only John came (see Ko¨nig and Siemund 1999; Eckardt 2006). Focus particles implicate alternative values, but restrict the head to just one of these. Among “intensifiers” that meet this criterion, self has received much attention. In He himself praised me/He praised me himself, himself focuses or excludes others than the referent (in this case the subject) from consideration, In English the reflexive (He praised himself) is derived from this intensifier use. The distinction between the focus and reflexive -self began to be worked out in Middle English (see Vezzosi 2007 for pragmatic arguments concerning this development). Prior to Middle English the reflexive was expressed by the unmarked personal pronoun. Vezzosi (2007: 253) argues that “Because of its meaning as an intensifier, himself in argument position could be stylistically exploited to convey authors’ subjective judgements or opinions on the action and/or actors”. In Middle English it was also used “to establish the topic of a narrative episode, […] to signal a new or unexpected topic, […] to characterize action as undesirable or condemnable” (250). She cites (5) to show the contrast with the neutral alternative him. In (5), hine seolf can only be understood as identifying Arthur: (5)
He [Baldolf] hafde iþohte bi nihten mid seouen þusend cnihten To riden uppen Arður ær þe king wore war And his folc afeollen and hine seolf aquelle (c.1200 Layamon Brut 870–873; Vezzosi 2007: 251) ‘He had thought he would come upon Arthur at night with 7000 knights before the king realized it, destroy his army and kill him himself’ [not kill himself!]
5 Pragmatic markers The class of pragmatic markers or “discourse particles” (Fischer 2006) is large and has been classified in many ways. According to Fraser (2006: 189) “These expressions occur as part of a discourse segment but are not part of the propositional content of the message conveyed, and they do not contribute to the meaning of the proposition per se”. The foundations of research on the historical development of pragmatic markers in English were laid in Brinton (1996). Here particular attention is paid to expressions that perform “salient event marking (i.e. foregrounding) and narrative segmentation (i.e. episode-boundary marking)” (268–269). In Middle English such expressions include bifel ‘it happened’, gan ‘began’, whan ‘when’, anon ‘at once’, and such epistemic parentheticals as I gesse ‘I guess’. Episode-marking with the impersonal bifel, which persisted from Old English, marked main clauses, and served an orienting function.
30 Middle English: Pragmatics and discourse It was replaced in later Middle English by episode-marking of subordinate clauses with whan ‘when’, which served a primarily framing function. This is seen as symptomatic of a redistribution of discourse functions associated with syntactic reorganization (179, 278). Gan is shown to be more foregrounding than anon, and I gesse is analyzed as a hedge, with strongly interpersonal function. A larger repertoire of pragmatic markers is discussed in Brinton (2006). In addition to the markers mentioned above, the following are discussed with examples from ME: whilom ‘once upon a time’, methinketh ‘it seems to me’, and the epistemic parenthetical (þæs/as) I wene ‘as I think’. I think is among the most extensively studied pragmatic markers in English. It competed with and eventually replaced the impersonal epistemic expression, methinketh/ methinks ‘it seems to me’ (Palander-Collin 1999). Brinton (1996) argues that epistemic parentheticals originated not in main clauses, but in as-parentheticals like ME as I guess, as it thynketh me ‘as seems to me’. Comment clauses more generally are the topic of Brinton (2008). Here she shows that say ‘seeking response/attention’ is one of the few that developed in earlier Middle English. Most comment clauses developed in late Middle English, including as it were, as you see, I dare say, I say ‘expression of emphasis’. Brinton concludes that construction grammar is a particularly appropriate framework for the analysis of comment clauses as it captures the alignment of individual instances of grammaticalized clauses with sets of clauses with similar properties (Brinton 2008: 256). Adverbs that originated in epistemic manner adverbs and came to be used in Middle English as intensifiers, and then as signals of the global structure of new episodes, are discussed in Lenker (2007). These include certes ‘certainly’, certeyn(ly), forsoothe ‘truly’, iwis ‘truly’, soothly ‘truly’, and trewely (82). Drawing on Grice’s (1989 [1975]) maxims and invited inferences, Lenker proposes that overuse of epistemic adverbials, especially such as occurs in clause-initial position, violates the maxims of quality (‘Do not talk about the veracity of the proposition unless there is some doubt about it’), and quantity (‘Do not make your contribution more informative than is required’). However, cooperation is preserved because invited inferences suggest that “truth or factuality has to be found on another than the propositional or lexical level”, i.e. at the level of textual organization (Lenker 2007: 100). Reformulation markers are investigated with respect to their functions as well as their forms in Late Middle English in Pahta and Nevanlinna (1997). While and and or have consistently been used in most text-types from earliest times, markers like that is (to say), namely are fairly common only in formal or technical writing (168). Their main function was paraphrase, and explanation of forms that are foreign or dialectal.
6 Speech acts There has been considerable discussion of methodologies for diachronic speech-act analysis in general (e.g. Arnovick 2006; Kohnen 2008), and for identifying the kinds of discursive traditions revealed by medieval texts. As part of a larger project on the history of speech acts in the history of English (Jucker and Taavitsainen 2008), Jucker (2000) investigates forms of verbal aggression in Chaucer’s Canterbury Tales. He points out that Chaucer’s insults intersect with two strands of research: on the one hand, profanity and oaths, which are associated with religious discourse, and on the other hand, flyting,
471
472
IV Middle English or ritualized verbal dueling, which is associated with secular interlocutors, such as knights. While slander is potentially truth-conditional, insults are not. In the Canterbury Tales insults are shown to be more personal than flyting, and more dangerous to interpersonal relationships. The article addresses the way in which such speech acts can be used as fictional devices that occur at several levels within the complex framing of the Tales. The levels can be classified as direct (e.g. the Miller insults the Reeve, the Host the Pardoner, etc. in the external frame), embedded (e.g. Palamon insults Arcite in the Knight’s Tale, as part of the narrative), mediated (e.g. narrative-internal insulting in the Friar’s Tale is intended to affect the Summoner), and indirect (e.g. the Friar addresses all pilgrims and introduces his tale of a Summoner in a way that indirectly insults the pilgrim Summoner). While cursing and swearing are relatively easy to identify in expressions such as by God, by Jhesus Christ, insults are harder to identify. Jucker proposes that reactions by the target appear to be the best cues to insults. Rudanko (2004) analyzes the role of threats in Chaucer’s wooing scenes, and suggests that speech act theory helps articulate the complexity of ways in which consent was perceived at the time. In another study of how Chaucer manipulates a particular type of speech act, in this case the “book curse”, Arnovick (2000) shows how he coopts a speech act for metalinguistic purposes in The House of Fame. The “book curse” is a common-place of medieval texts, usually designed to claim ownership and curse anyone who does not treat the book well, e.g. by desecrating it, or copying it poorly. In The House of Fame it is used, Arnovick argues, to confer ironic import, help the audience suspend disbelief, and project themselves into Chaucer’s dream world. Pakkala-Weckstro¨m (2002) addresses the critical problem of how to understand Dorigen’s “rash promise” in Chaucer’s Franklin’s Tale, a promise that almost destroys her, her husband, and her would-be lover Aurelius. Pakkala-Weckstro¨m examines Aurelius’s failure to live up to the standard of the courtly lover, and suggests that Dorigen’s speech acts are trumped by her husband’s honor and the value of trouthe ‘solemn promise’, as understood in the context of the genre of the romance. By contrast, Arnovick suggests that the Tale is about “interpretation” in “three oral realms – linguistic, legal, and folk” (Arnovick 2006: 173–174). While the studies mentioned above address highly vocal expressiveness, Arnovick (2006) explores the discursive function of silence in Biblical tradition and monastic practice, and Hiltunen (2003) and Green (2007) draw attention to practices of silence versus institutionally sanctioned talk in both spiritual and secular contexts. Hiltunen interprets instructions to the anchoresses in Sawles Warde in terms of Gricean conversational maxims. Green, on the other hand, questions the appropriateness of conversational maxims for institutional and group settings. These and other works on speech acts are neophilological in orientation, combining linguistic analysis with textual interpretation. We may note that much work on speech acts in Middle English has drawn on Grice’s maxims, which highlight logic and cooperation in conversation. To the extent that these acts are aggressive, contesting, and impolite, they flout these maxims, or are more directly related to Keller’s (1994 [1990]) dynamic maxims, which concern social interaction and identity formation. However, to date, the latter do not appear to have impacted work on Middle English directly.
30 Middle English: Pragmatics and discourse
7 Politeness strategies Speech acts are inextricably intertwined with issues of politeness. Other linguistic domains that intersect with politeness include pragmatic markers (I guess), nominal forms of address (lord, barn ‘child, youth’, caytif ‘wretch, scoundrel’), and second person pronouns (þu, ye with singular reference [T/V forms]). The first of these is discussed in e.g. Brinton (1996) (see Section 5), the second in e.g. Honegger (2005), and the third in e.g. Blake (1992). Blake (1992: 537) cautions against too facile acceptance of Skeat’s (1867: xlii) claim that “thou is the language of a lord to a servant, of an equal to an equal, and expresses also companionship, love, permission, defiance, scorn, threatening; whilst ye is the language of a servant to a lord, and of compliment, and further expresses honour, submission, entreaty”. In other words Blake argues against too strict a view of a “power semantic” (Brown and Gilman 1960) in the use of T/V pronouns, as does Burnley (2003). The extent to which ye was used to signal social distance may vary according to genre (romances, for example, favor ye), author, or text. Burnley discusses the rejection of the T-form by certain religious groups, particularly in Europe, and warns that because we do not know to what extent Chaucer was aware of or cared about humanistic debates on address forms in Italy and France, we cannot fully assess his usage with certainty. More positively, Jucker (2006) argues that Chaucer uses address forms in ways that show social status and the characters’ relationships to each other, but primarily they index momentary status attained in the course of interactional negotiation. If thou was favored in certain religious groups, by the later ME period, ye/you appears to have become the norm at least among the middle class. Bergs (2004) shows that there are no uses of T-forms in the Stonor Letters (1290– 1483) or the Plumpton Letters (1433–1551). Only a scattering is found in the Cely Letters (1472–1488), and only in letters by the elder Cely to his son when the latter was young; under 1% is found in the Paston Letters (1421–1503). Bergs (2004: 134–135) suggests that use of V-pronouns may be a characteristic of letters as a text-type in Middle English. (See also Jucker, Chapter 13.)
8 Pragmatic and discourse properties of genres and text-types Here “genre” is understood to refer to macro-level characterizations of text-types, e.g. romance, drama, and “text-type” to more specific sub-categorizations of genre (however, genre and text-type are often used interchangeably in the literature on Middle English). For discussion of the variety of ways in which the terms genre, text-type and register (see Section 9) are used, see Diller (2001). Drawing attention to the fact that some text-type labels may refer to very different text-types across time, Taavitsainen (1993) shows how a set of features such as “affective mood” can be used to identify differences among subgenres and styles in later medieval texts, and to test subgenre classification in corpora such as the Helsinki Corpus (Rissanen et al. 1991). “Affective” forms include personal pronouns, question-words, private verbs of emotion and states of perception. Statistical analysis is used to show that in the romances affective features appear primarily in direct speech. While this may hardly be surprising, it serves as a methodological caution: text-types may be internally highly diverse, or overlap: “The only reliable starting point for research on Middle English genres is with the individual texts themselves”
473
474
IV Middle English (Taavitsainen 1993: 195). Kohnen (2001: 122) suggests that the period 1300–1600 “may be called the period of text types” in that one can trace the facilitation of the spread of specific constructions such as Latin-based adverbial uses of the –ing participle as in (6) across text-types as they became vernacularized: (6)
he sawe soules hange þerin, crying, wayling and morning for wo, peyn and sorowe (Advent and Nativity Sermons [Powell 1981]: 78; Kohnen 2001: 118) ‘he saw souls hang therein, crying, wailing and mourning for woe, pain, and sorrow’
Drawing on Haugen’s (1972) distinction between elaboration and codification in standardization, he proposes further that text type-dependent development should be regarded as elaboration; codification comes with text-independence in the later period (Kohnen 2001: 121–122). Among text-types discussed in some detail in recent work with respect to their distinctive discourse characteristics are instructional “how to” documents (Carroll 2003), medical recipes (Almeida and Carroll 2004; Ma¨kinen 2004), wills (Del Lungo Camiciotti 2002), and catalogs of tracts, doctrinal disputations, and statutes (Peikola 2003). Vernacularization of scientific texts, modeling on Latin and Arabic, and the professionalization of English are among topics discussed in Taavitsainen and Pahta (2004), and metalinguistic conventions for letters in Nevalainen (2001). Del Lungo Camiciotti (2007) addresses the refashioning of saints’ lives from hagiographic text to narrative in the late 14th and 15th centuries, using the example of Bokenham’s Legendys of Hooly Wummen (1447). She shows how represented dialog is used in ways analogous to contemporary genres such as devotional treatises and drama (2007: 288) to structure the narrative and at the same time “create involvement in narrated events by dramatizing” (2007: 297). Some studies involving text-types start with particular linguistic constructions, or sets of constructions, and discuss the extent to which they are used in specific text-types. For example, basing her analysis on the HC and other corpora, Kyto¨ (1997) details the gradual increase of the use of perfect have + past participle at the expense of be + past participle with verbs of transition and motion like come, go, arrive until the late 1700s when have came to predominate. The main linguistic factors are shown to include “action/ process uses, durative, iterative and conditional verb contexts” as well as presence of a complement (70). Text-type factors in Late Middle English show that the have-perfect was used on a continuum extending from highest in fiction (40%), through diaries (37%), sermons (24%), private letters (10%), to least in drama (10%), and zero in scientific writing. No significant differences were found regarding orality or informality. In their study of “central modals” from the late 14th to the late 17th century, Gotti et al. (2002) investigate which modals were most frequent in the text-types represented in HC. Their findings are that shall occurs with unusually high frequency in Rules, where the modal expresses obligation, and in religious treatises and Biblical texts, where it mainly expresses future, or God’s will. May and might are most frequent in non-private letters. Whereas philosophy shows an above-average use of may, narrative and history favor might. While may and might occur in primarily formal texts, can and could appear in more speech-based texts, such as dialogues in the Canterbury Tales, and sermons (330). Code-switching and “macaronic” usage, in which two or more languages are mixed in a fairly stylized way, occur in many ME texts. In Middle English, Latin was the “high”
30 Middle English: Pragmatics and discourse variety of religion and scholarship, French ranked at the middle level (until it became largely restricted to law in the 14th century), and English the “low” variety. Schendl (2001) exemplifies the structural characteristics of syntactic code-switching in ME poetry. Given the monologual nature of this poetry, he hypothesizes that the function of switching may have been practice in writing poetry, indexing membership in the educated class, and language play. Bergner (1998) associates the use of Latin in the mystery plays with aureate style, in part because it is limited to characters who consistently speak with authority, such as God. Where other text-types are concerned, it may play a different role. Pahta (2004) identifies code-switching in medical texts with specialized terminology, intertextuality (quotations), taboo topics, and formulaic tags such as Latin probatum est ‘it has been tested’ to attest to the value of recipes. Wright (2005) discusses the development of Anglo-French vocabulary and affixes in England, as exemplified by business letters, and tracks correlations between use of Anglo-French and centralization of commerce in London. She suggests that the development of Standard English in later centuries was a side-effect of these ME changes in patterns of commerce.
9 Register Since the vast majority of ME texts were intended to be read aloud as well as silently (e.g. sermons), or were dictated (e.g. letters by women), categories like “speech-based” (e.g. sermons, homilies) vs. “non-speech-based” (scientific texts, philosophical works, travelogues letters) are ultimately not very useful in Middle English (or earlier), even though texts in the HC (Rissanen et al. 1991) are coded for these factors. Many researchers have therefore turned to the oral-literate or informal-formal continuum (see e.g. Tannen 1982; Biber and Finegan 1997). However, because orality and literacy are inter-related prior to the crystallization of genre and text-types (Wa˚rvik 2003), the “continuum” should not be thought of as linear. Oral habits left a mark on the composition of texts, and the texts “are full of idiosyncrasies and incoherences, with gaps, anomalies of grammar, and incoherent discourse patterns” (Taavitsainen and Fitzmaurice 2007: 19) that do not appear in texts on either end of the oral-literate continuum, but rather suggest emergent practices (Arnovick 2006; Del Lungo Camiciotti 2007). Sikorska (2000) argues that oral and written discourses are interdependent in The Book of Margery Kempe (c.1450). Highlighting the tension between author, scribe, and audience that Kempe must have been trying to overcome when she composed her book, Sikorska hypothesizes that, being secular, Kempe used third person narrative to maintain distance and project modesty in order to legitimate her visions. These might be considered to be on the literate end of the continuum. However, Kempe also uses “oral” expressions such as interjections (A, good Lord) and discourse markers like gan and anon. Her use of whan ‘when’ is similar to that suggested for Middle English by Brinton (1996), i.e. text-organizing, but also should be understood as primarily iterative (‘whenever’); it suggests “general truth”, or “her recognition of a situation recently described” (Sikorska 2000: 404). In the later periods of English, drama is often thought of as a site where historians of English might find some evidence for colloquial or spoken language. However, as Culpeper and Kyto¨ (2000) show, speech is represented and filtered in drama even in the EModE period, and therefore cannot be taken as direct evidence for speech. In Middle English the conventions of drama do not favor realistic speech. Bergner (1998)
475
476
IV Middle English characterizes the speech in the mystery plays as being of two types. One is “vertical” speech spoken by God, Jesus, Mary and other characters of biblical authority. These characters give information, foretell, preach, interpret, and harmonize, often in monologues. The diction is “aureate”, with Latin code-switching or Latinate vocabulary. The other type is “horizontal”, spoken by persons of generally low birth, such as the shepherds and the soldiers. Their speech is short and appears spontaneous, marked by interjections, interruptions, and flyting. While Noah, his wife, and the shepherds engage in such speech in the early parts of the plays in which they appear, when Noah converses with God or when the shepherds arrive at the manger, they adopt a more vertical style. The two styles complement each other, “[o]ne form of speech being justified in its existence only through the presence of the other” (Bergner 1998: 82). Again, what are often considered literate and oral cues are interdependent.
10 Middle English as a period of transition Lass (2000) points to some of the absurdities of periodization, especially of the notion “Middle” when “Modern” (“our period”) is forever expanding. He does, however, conclude that “there are apparently suites of characters that define” Middle English (35). He mentions phonological and morphological factors, and word order. To these we might add some pragmatic and discoursal factors. Most notable is the gradual alignment of topic with subject and givenness (but not animacy, see Section 2) as the syntacticization of word order proceeded, specifically the shift toward a greater distinction between main and subordinate clauses (hypotaxis). At the same time, the forms and functions of markers of discourse relations underwent some significant changes (Section 5). Entrenchment in the system for the most part did not occur until the EModE period, followed by further modification, e.g. the association of animacy with topic. Practices and conventions of writing also changed throughout Middle English as new genres were borrowed and as increasing vernacularization occurred, whether of the Bible or of scientific texts.
11 References ´ lvarez, Alicia Rodrı´guez and Francisco Alonso Almeida (eds.). 2004. Voices on the Past: Studies A in Old and Middle English Language and Literature. Spain: Netbiblo, S. L. Almeida, Francisco Alonso and Ruth Carroll. 2004. A new proposal for the classification of Mid´ lvarez and Almeida (eds.), 21–33. dle English medical texts. In: A Arnovick, Leslie K. 2000. Whoso thorgh presumpcion … mysdeme hyt: Chaucer’s poetic adaptation of the medieval “book curse”. In: Taavitsainen et al. (eds.), 411–424. Arnovick, Leslie K. 2006. Written Reliquaries: The Resonance of Orality in Medieval English Texts. Amsterdam/Philadelphia: John Benjamins. Bergner, Heinz. 1998. Dialogue in the medieval drama. In: Raimund Borgmeier, Herbert Grabes, and Andreas H. Jucker (eds.), Anglistentag 1997 Giessen Proceedings, 75–83. Trier: WVT Wissenschaftlicher Verlag. ´ lvarez and Almeida Bergs, Alexander T. 2004. Address pronouns in Late Middle English. In: A (eds.), 127–138. Biber, Douglas and Edward Finegan. 1997. The linguistic evolution of five written and speechbased English genres from the 17th to the 20th centuries. In: Matti Rissanen, Ossi Ihalainen, Terttu Nevalainen, and Irma Taavitsainen (eds.), History of Englishes: New Methods and Interpretations in Historical Linguistics, 688–704. Berlin/New York: Mouton de Gruyter.
30 Middle English: Pragmatics and discourse Blake, Norman. 1992. The literary language. In: Norman Blake (ed.), The Cambridge History of the English Language. Vol. II. 1066-1476, 500–540. Cambridge: Cambridge University Press. Brinton, Laurel J. 1996. Pragmatic Markers in English: Grammaticalization and Discourse Functions. Berlin/New York: Mouton de Gruyter. Brinton, Laurel. 2006. Pathways in the development of pragmatic markers in English. In: van Kemenade and Los (eds.), 307–334. Brinton, Laurel J. 2008. The Comment Clause in English: Syntactic Origins and Pragmatic Development. Cambridge: Cambridge University Press. Brown, Roger and Albert Gilman. 1960. Pronouns of power and solidarity. In: Thomas Sebeok (ed.), Style in Language, 253–276. Cambridge, MA: MIT Press. Burnley, David. 2003. The T/V pronouns in later Middle English. In: Irma Taavitsainen and Andreas H. Jucker (eds.), Diachronic Perspectives on Address Term Systems, 27–45. Amsterdam/ Philadelphia: John Benjamins. Carroll, Ruth. 2003. Recipes for laces: An example of a Middle English discourse colony. In: Hiltunen and Skaffari (eds.), 137–165. Culpeper, Jonathan and Merja Kyto¨. 2000. Data in historical pragmatics: Spoken interaction (re) cast as writing. Journal of Historical Pragmatics 1: 175–199. Danchev, Andrei and Merja Kyto¨. 1994. The construction be going to + infinitive in Early Modern English. In: Kastovsky (ed.), 59–77. Davis, Norman. 1971. Paston Letters and Papers of the Fifteenth Century. Vol. XX. Oxford: Clarendon Press. Del Lungo Camiciotti, Gabriella. 2002. Performative aspects of late medieval English wills. Journal of Historical Pragmatics 3: 205–227. Del Lungo Camiciotti, Gabriella. 2007. Discoursal aspects of the Legends of Holy Women by Osbern Bokenham. In: Fitzmaurice and Taavitsainen (eds.), 285–305. Diller, Hans-Ju¨rgen. 2001. Genre in linguistic and related discourses. In: Diller and Go¨rlach (eds.), 3–43. Diller, Hans-Ju¨rgen and Manfred Go¨rlach (eds.). 2001. Towards a History of English as a History of Genres. Heidelberg: Winter. Eckardt, Regine. 2006. Meaning Change in Grammaticalization. Oxford/New York: Oxford University Press. Evans, Nicholas and David Wilkins. 2000. In the mind’s ear: The semantic extensions of perception verbs in Australian languages. Language 76: 546–592. Fischer, Olga and Wim van der Wurff. 2006. Syntax. In: Richard Hogg and David Denison (eds.), A History of the English Language, 109–198. Cambridge: Cambridge University Press. Fitzmaurice, Susan and Irma Taavitsainen (eds.). 2007. Methods in Historical Pragmatics. Berlin/ New York: Mouton de Gruyter. Fraser, Bruce. 2006. Towards a theory of discourse markers. In: Kerstin Fischer (ed.), Approaches to Discourse Particles, 189–204. Amsterdam/Oxford: Elsevier. Gotti, Maurizio. 2005. Prediction in Middle English: A comparison between shall and will. In: Ritt and Schendl (eds.), 196–215. Gotti, Maurizio, Marina Dossena, Richard Dury, Roberta Facchinetti, and Maria Lima (eds.). 2002. Variation in Central Modals: A Repertoire of Forms and Types of Usage in Middle English and Early Modern English. Bern: Lang. Green, Eugene. 2007. Patterns of seeking and breaking silence in earlier English. In: Christopher M. Cain and Geoffrey Russom (eds.), Studies in the History of the English Language III. Managing Chaos: Strategies for Identifying Change in English, 215–229. Berlin/New York: Mouton de Gruyter. Grice, H. Paul. 1989 [1975]. Logic and conversation. Studies in the Way of Words, 22–40. Cambridge, MA/London: Harvard University Press. Haugen, Einar. 1972. Dialect, language, and nation. In: J. B. Pride and Janet Holmes (eds.), Sociolinguistics: Selected Readings, 97–111. Harmondsworth: Penguin.
477
478
IV Middle English Hiltunen, Risto. 2003. Telling the anchorite code: Ancrene Wisse on language. In: Hiltunen and Skaffari (eds.), 57–76. Hiltunen, Risto and Janne Skaffari (eds.). 2003. Discourse Perspectives on English. Amsterdam/ Philadelphia: John Benjamins. Honegger, Thomas. 2005. ‘Wyȝe welcum iwys to this place’ – and never mind the alliteration: An inquiry into the use and forms of address in two alliterative ME romances. In: Ritt and Schendl (eds.), 169–178. Jucker, Andreas H. 2000. Slanders, slurs, and insults on the road to Canterbury: Forms of verbal aggression in Chaucer’s Canterbury Tales. In: Taavitsainen et al. (eds.), 369–389. Jucker, Andreas H. 2006. “Thou art so loothly and so oold also”: The use of ye and thou in Chaucer’s Canterbury Tales. Anglistik 17: 57–72. Jucker, Andreas H. and Irma Taavitsainen (eds.). 2008. Speech Acts in the History of English. Amsterdam/Philadelphia: John Benjamins. Kastovsky, Dieter (ed.). 1994. Studies in Early Modern English. Berlin/New York: Mouton de Gruyter. Kemenade, Ans. van 2009. Discourse relations and word order change. In: Roland Hinterho¨lzl and Svetlana Petrova (eds.), Information Structure and Language Change: New Approaches to Word Order Variation and Change in the Germanic Languages, 91–118. Berlin/New York: Mouton de Gruyter. Kemenade, Ans. van and Bettelou Los. 2006. Discourse adverbs and clausal syntax in Old and Middle English. In: van Kemenade and Los (eds.), 224–246. Kemenade, Ans. van and Bettelou Los (eds.). 2006. Handbook of the History of English. Oxford: Blackwell. Keller, Rudi. 1994 [1990]. On Language Change: The Invisible Hand in Language. Brigitte Nerlich (trans.). London: Routledge. Kohnen, Thomas. 2001. Text types as catalysts for language change: The example of the adverbial first participle construction. In: Diller and Go¨rlach (eds.), 111–124. Kohnen, Thomas. 2008. Tracing directives through text and time: Towards a methodology of corpus-based diachronic speech-act analysis. In: Jucker and Taavitsainen (eds.), 295–310. Ko¨nig, Ekkehard and Peter Siemund. 1999. Intensifiers as targets and sources of semantic change. In: Andreas Blank and Peter Koch (eds.), Historical Semantics and Cognition, 237–257. Berlin/ New York: Mouton de Gruyter. Krug, Manfred G. 2002. Emerging English Modals: A Corpus-based Study of Grammaticalization. Berlin/New York: Mouton de Gruyter. Kyto¨, Merja. 1997. Be/have + past participle: The choice of the auxiliary with intransitives from Late Middle to Modern English. In: Rissanen, Kyto¨, and Heikkonen (eds.), 17–85. Kyto¨, Merja and Suzanne Romaine. 2005. We had like to have been killed by thunder and lightning: The semantic and pragmatic history of a construction that like to disappeared. Journal of Historical Pragmatics 6: 1–35. Lass, Roger. 2000. Language periodization and the concept “middle”. In: Taavitsainen et al. (eds.), 7–41. Lenker, Ursula. 2007. Soþlice, forsoothe, truly – communicative principles and invited inferences in the history of truth-intensifying adverbs in English. In: Fitzmaurice and Taavitsainen (eds.), 81–105. Los, Bettelou. 2005. The Rise of the to-Infinitive. Oxford: Oxford University Press. Mubashshir ibn Fatik, Abu al-Wafa’. 1477. Dictes or sayengis of the philosophhres. Early English Books Online. See eebo.chadwyck.com/ Ma¨kinen, Martti. 2004. Herbal recipes and recipes in herbals – intertextuality in early English medical writing. In: Taavitsainen and Pahta (eds.), 144–173. Mazzon, Gabriella (ed.). 2007. Studies in Middle English Forms and Meanings. Frankfurt: Lang. Me´ndez-Naya, Bele´n. 2007. He nas nat right fat: On the origin and development of the intensifier right. In: Mazzon (ed.), 191–207. Nevalainen, Terttu, Helena Raumolin-Brunberg, Jukka Kera¨nen, Minna Nevala, Arja Nurmi, and Minna Palander-Collin. 1998. Corpus of Early English Correspondence (CEEC). Department of English, University of Helsinki. http://www.helsinki.fi/varieng/CoRD/corpora/CEEC/index.html
30 Middle English: Pragmatics and discourse Nevalainen, Terttu. 1994. Aspects of adverbial change in Early Modern English. In: Kastovsky (ed.), 359–380. Nevalainen, Terttu. 2001. Continental conventions in early English correspondence. In: Diller and Go¨rlach (eds.), 203–224. Pahta, Pa¨ivi. 2004. Code-switching in medieval medical writing. In: Taavitsainen and Pahta (eds.), 73–99. Pahta, Pa¨ivi and Saara Nevanlinna. 1997. Rephrasing in Early English: The use of expository apposition with an explicit marker from 1350 to 1710. In: Rissanen, Kyto¨, and Heikkonen (eds.), 121–183. Pakkala-Weckstro¨m, Mari. 2002. Have her my trouthe – Til that myn herte breste: Dorigen and the difficulty of keeping promises in the Franklin’s Tale. In: Helena Raumolin-Brunberg, Minna Nevala, Arja Nurmi, and Matti Risannen (eds.), Variation Past and Present: VARIENG Studies on English for Terttu Nevalainen, 287–300. Helsinki: Socie´te´ Ne´ophilologique. Palander-Collin, Minna. 1999. Grammaticalization and Social Embedding: I THINK and METHINKS in Middle and Early Modern English. Helsinki: Socie´te´ Ne´ophilologique. Paradis, Carita. 1997. Degree Modifiers of Adjectives in Spoken British English. Lund: Lund University Press. Paradis, Carita. 2000. It’s well weird: Degree modifiers of adjectives revisited: The nineties. In: John M. Kirk (ed.), Corpora Galore: Analyses and Techniques in Describing English: Papers from the Nineteenth International Conference on English Language Research on Computational Corpora (ICAME 1988), 146–160. Amsterdam: Rodopi. Peikola, Matti. 2003. The catalogue: A late Middle English Lollard genre? In: Hiltunen and Skaffari (eds.), 105–135. Peters, Hans. 1994. Degree adverbs in Early Modern English. In: Kastovsky (ed.), 269–288. Powell, Susan (ed.). 1981. The Advent and Nativity Sermons from a Fifteenth-Century Revision of John Mirk’s Festial. Heidelberg: Winter. Rissanen, Matti, Merja Kyto¨, Leena Kahlas-Tarkka, Matti Kilpio¨, Saara Nevanlinna, Irma Taavitsainen, Terttu Nevalainen, and Helena Raumolin-Brunberg. 1991. The Helsinki Corpus of English Texts. In: ICAME Collection of English Language Corpora (CD-ROM), 2nd edn., eds. Knut Hofland, Anne Lindebjerg, and Jørn Thunestvedt, The HIT Centre, University of Bergen, Norway. For manual, see http://khnt.hit.uib.no/icame/manuals/hc/index.htm Rissanen, Matti, Merja Kyto¨, and Kirsi Heikkonen (eds.). 1997. English in Transition: Corpusbased Studies in Linguistic Variation and Genre Styles. Berlin/New York: Mouton de Gruyter. Ritt, Nikolaus and Herbert Schendl (eds.). 2005. Rethinking Middle English: Linguistic and Literary Approaches. Frankfurt am Main: Lang. Rudanko, Juhani. 2004. “I wol sterve”: Negotiating the issue of a lady’s consent in Chaucer’s poetry. Journal of Historical Pragmatics 5: 137–158. Schiffrin, Deborah. 1992. Anaphoric then: Aspectual, textual, and epistemic meaning. Linguistics 30: 753–792. Schendl, Herbert. 2001. Code-switching in medieval English poetry. In: Dieter Kastovsky and Arthur Mettinger (eds.), Language Contact in the History of English, 305–335. Frankfurt: Lang. Seoane, Elena. 2006. Information structure and word order change: The passive as an informationrearranging strategy in the history of English. In: van Kemenade and Los (eds.), 360–391. Sikorska, Liliana. 2000. Hir not lettyred: The use of interjections, pragmatic markers and whanclauses in The Book of Margery Kempe. In: Taavitsainen et al. (eds.), 392–410. Skeat, Walter W. 1867. William of Palerne: To which is Added a Fragment of the Alliterative Romance of Alexander. London: Bungay. Stoffel, C. 1901. Intensives and Down-Toners: A Study in English Adverbs. Heidelberg: Winter. Stubbs, Michael. 1995. Collocations and semantic profiles: On the cause of trouble with quantitative studies. Functions of Language 2: 23–56.
479
480
IV Middle English Taavitsainen, Irma. 1993. Genre/subgenre style in Late Middle English. In: Matti Rissanen, Merja Kyto¨, and Minna Palander-Collin (eds.), Early English in the Computer Age: Explorations through the Helsinki Corpus, 171–200. Berlin/New York: Mouton de Gruyter. Taavitsainen, Irma and Susan Fitzmaurice. 2007. Historical pragmatics: What it is and how to do it. In: Fitzmaurice and Taavitsainen (eds.), 11–36. Taavitsainen, Irma, Terttu Nevalainen, Pa¨ivi Pahta, and Matti Rissanen (eds.). 2000. Placing Middle English in Context. Berlin/New York: Mouton de Gruyter. Taavitsainen, Irma and Pa¨ivi Pahta (eds.). 2004. Medical and Scientific Writing in Late Medieval English. Cambridge: Cambridge University Press. Tannen, Deborah. 1982. The oral/literate continuum in discourse. In Deborah Tannen (ed.), Spoken and Written Language, 1–16. Norwood, NJ: Ablex. Taylor, Ann and Wim van der Wurff (eds.). 2005. Aspects of OV and VO word order in the history of English. Special issue of English Language and Linguistics 9. Traugott, Elizabeth Closs. 2006. The semantic development of scalar focus modifiers. In: Kemenade and Los (eds.), 335–359. Vezzosi, Letizia. 2007. Himself: An overview of its use in ME. In: Mazzon (ed.), 239–256. Wa˚rvik, Brita. 2003. “When you read or hear this story read”: Issues of orality and literacy in Old English texts. In: Hiltunen and Skaffari (eds.), 113–155. Wright, Laura. 2005. Medieval mixed-language business discourse and the rise of Standard English. In: Janne Skaffari, Matti Peikola, Ruth Carroll, Risto Hiltunen, and Brita Wa˚rvik (eds.), Opening Windows on Texts and Discourses of the Past, 381–399. Amsterdam/Philadelphia: Benjamins.
Elizabeth Closs Traugott, Palo Alto (USA)
31 Middle English: Dialects 1 2 3 4 5 6 7 8
Introduction Previous conceptions of dialectal variation in Middle English A spatio-temporal frame for Middle English “dialects” Reconstructing the dialect continuum for Middle English Dialectal variation in Middle English Dialects and word geography Further prospects References
Abstract This chapter offers a conspectus of the kinds of regional variation to be inferred from the study of the language of Middle English texts. The perspective is that which underlies A Linguistic Atlas of Late Middle English, 1350–1450 (McIntosh et al. 1986) and A Linguistic Atlas of Early Middle English (Laing and Lass 2008). The “traditional” view that there were such things as dialects in Middle English is eschewed in favor of recognizing the “dialectology” of Middle English as a continuum of overlapping feature distributions. The methodology of this approach is briefly outlined. Examples of a set of significant distributions of linguistic features are illustrated cartographically and discussed. Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 480–505
31 Middle English: Dialects
1 Introduction The purpose of this chapter is to provide an overview of the kind of regionally marked variation to be derived from analysis and interpretation of the language of Middle English (ME) texts. The title of the chapter is “Dialects”, but this term requires clarification. The most significant and paradoxical finding of dialect geography has been the non-existence of dialects, the objects which it purports to study. A “dialect” is a construct: a reification of some assemblage of linguistic features, defined according to criteria established by the dialectologist. These criteria may be linguistic, extralinguistic, or some combination of these two kinds. (For more detailed discussion see Benskin 1994.) Identification of dialects is premised (at least partly) on the interpretation of linguistic variation distributed across geographical space. “Middle English” has traditionally been defined as a set of “dialects” in the sense that the language is characterized by a high degree of variation, and a considerable number of linguistic variants can be identified as regionally marked. The issue is how this regional variation can best be determined and understood given the complex nature of the extant data: an accidentally surviving corpus of manuscript texts, dating from the mid-12th century to the mid-15th century about whose producers we have little or no direct knowledge or information in most cases. If we want to have a realistic categorical conception of “dialect” for Middle English, we can consider each scribal text as one. However, in many cases even that would be too simple a characterization of the language of the text (see further Benskin and Laing 1981; Laing 2004; Laing and Williamson 2004). A (Middle English) dialect can be considered to be some assemblage of diatopically coherent linguistic features which co-occur over all or part of their geographical distributions and so delineate an area within the dialect continuum. Adding or taking away a feature from the assemblage is likely to alter the shape of this area: addition might reduce the area’s size, subtraction, increase it. Any text which contains that assemblage of features can be considered as having its provenance within that area. Linguistic variation across space exists because languages undergo change as a natural consequence of use and transmission from generation to generation of speakers. A linguistic change takes place in a community and subsequently becomes disseminated through time and across space as the next generation of speakers and neighboring communities adopt the change. A “dialect continuum” emerges from the language contacts of speakers in neighboring communities as they share some features, but not others. A continuum is thus made up of a set of overlapping distributions of linguistic features with varied geographical extents – some extensive, some local. A continuum is not static and is constantly shifting more or less rapidly through time. The primary data for Middle English is written language. From the Norman Conquest of England until the middle of the 12th century, there are few surviving original compositions written in English. When written English begins to appear again, it is being reinvented. This reinvention takes place locally. New orthographic systems are formulated with greater or lesser degrees of consistency and economy to reflect, at least in part, regional, local, and even individual spoken varieties. And, as will be seen in the evidence discussed later, written symbols themselves may exhibit regional variation independent of reference to that of underlying spoken forms.
481
482
IV Middle English
2 Previous conceptions of dialectal variation in Middle English The earliest systematic attempts at defining ME dialects were the studies of Oakden (1930) and Moore et al. (1935). Oakden’s study was driven by an interest in the localization of the original manuscript versions of certain alliterative poems. It was also partly an attempt to improve on earlier work by Serjeantson (1927) to define what the characteristic linguistic features of the West Midlands were. Oakden has a list of linguistic “points” which he describes “of some dialectal value in localising texts, and particularly those of the fourteenth century”. His points are based on etymological characters and lump together a number of words or morphemes under a single “point”. His map shows isoglosses to indicate “the approximate boundaries of these features […] The map may then be used as a key for localising a document of a particular period with a considerable degree of accuracy” (Oakden 1930: 9). Moore et al. (1935) arrived at a basic conception of ten regional varieties of Middle English: Kentish, Southern, South-east Midland, North-east Midland, Southwest Midland, South-central West Midland, North-central West Midland, Northwest-Midland, Central East Midland and Northern. These were defined according to eleven discriminants. These divisions provided a basic architecture for the dialect map of England and underpinned the study of linguistic variation in Middle English for many years. Both of these studies present their information with a categorical cleanness that hardly begins to capture the complexity of the linguistic variation to be found in the linguistic witnesses – the texts. Gillis Kristensson’s linguistic-geographical studies of the largely onomastically-coded vernacular forms in the Lay Subsidy Rolls have provided a valuable further insight into the diatopic diversity of Middle English (Kristensson 1967, 1987, 1995, 2001a, 2001b) and present a much more nuanced picture, albeit the tendency to map etymological categories and to define boundaries of distributions by isoglosses results in a degree of sanitization of the data. There are also problems in the reliability of his sources – London-made copies of local texts – and in how much the names that Kristensson used for his data might have been altered by the copyists and perhaps not actually representative of local forms. (On this see McClure 1973. For citations of reviews of Kristensson 1967, see Kristensson 1987: ix–x, n. 2, and for his responses, p. ix seq.) Yet, none of the above mentioned studies offer any representation of language on a map that could be described as a dialect per se. Rather, however much the authors try to define boundaries, what emerges is a set of overlapping geographical distributions of linguistic characteristics. With the publication of A Linguistic Atlas of Late Mediaeval English, 1350–1450 (McIntosh et al. 1986, henceforth LALME) and of A Linguistic Atlas of Early Middle English, 1150–1325 (Laing and Lass 2008, henceforth LAEME) the complexity of linguistic variation throughout the Middle English period has been made apparent in unprecedented detail. Both these works have taken radically different approaches to the study of the dialectology of Middle English from those of the previous studies mentioned.
3 A spatio-temporal frame for Middle English “dialects” The conception of Middle English dialects offered here is informed largely by the methodology used to create LALME and LAEME and the results published in these
31 Middle English: Dialects enquiries. The geographical bounds for our study may be defined as England in the late Middle Ages (albeit its northern bounds with Scotland and its western bounds with Wales were not entirely static in this period). However, ME “dialects” are not solely a diatopic matter. There is also a crucial diachronic dimension. Quite when Middle English began and ended has been a matter of debate among historians of English (for an interesting discussion, see Kitson 1997). In particular, the question of defining a temporal boundary between “Middle” as opposed to “Old” English has generated much discussion. (For discussion of the notion “Middle” and references, see Lass 2000.) This matter is further complicated by the lack of surviving texts composed in English of the mid-11th century to the mid-12th century as opposed to later copies of Old English texts. The opposition between Old English and Middle English has been portrayed as one between language types: Old English is characterized by a more complex morphology and flexible word-order; Middle English is characterized by loss of most of the declensional and (to a lesser extent) conjugational morphology and by more rigid word-order; also, the lexis of Middle English contains a heavy borrowing of words from Old French following the Conquest, replacing or bringing about the re-referencing of their Old English equivalents. In fact, Old English was well on the way to a simpler morphological structure by the 11th century. What emerges as Middle English in the extant texts of the 12th century (lexical disturbance aside) is a continuation of this process. At the other end of the ME period, marked linguistic variation in the written language is giving way to an emergent normative variety which is spreading geographically through the mid and later 15th century. Regional diversity in the written language is smoothed out, first in the South and Midlands and later in the North. Of course, languages are constantly in transition and in examining the “dialects” of Middle English we are examining one period of transition defined by changes in the nature of surviving textual artefacts: texts which are demonstrably not-Old English at one end and texts which are demonstrably not-Standard(ized) at the other end. Through the ME period, linguistic change is observable in and deducible from the evidence of the texts. The direct sources of the linguistic information for Middle English are provided by the written forms found in the extant texts. For the spoken language, we can infer only from the evidence of the orthography of the texts and from reconstructive techniques based on the direct evidence of speech in the modern period. The regional variants which are characteristic of Middle English did not arise ex nihilo. And, if they largely vanish from the written record in the Early Modern period, they (and their developments) did not do so from the spoken language. The regional variation of the Late Modern period (as recorded notably in Alexander Ellis’s [1869–89] On Early English Pronunciation, Vol. V, and in Joseph Wright’s [1898–1905] English Dialect Dictionary for the 19th century, and in the Survey of English Dialects for the 20th century [Orton et al. 1962–71]) are the later developments of the spoken regional varieties of Middle English. See Dietz (1989) for a comparison of the distribution of linguistic features between the medieval and modern periods. That Old English was more varied in writing than the West Saxon literary texts of the 10th and 11th centuries would suggest is evident from the English to be found in charters, whether contemporary, or later copies preserving the Old English text, as Peter Kitson has demonstrated (Kitson 1993, 1995). For present purposes, the matter is not so much when and for what reasons one may wish to argue that Middle English began and ended, as the fact that
483
484
IV Middle English the period in which the forms found in surviving texts exhibit significant regional variation lasted several hundred years. It will suffice for present purposes to take as the period for discussion here from 1150 to 1475. We have, then, to deal with a period of 325 years. An historical dialectological approach must take into account such time-depth: a proper description of ME regional variation cannot be a synchronic one. In practice, a time-depth has been recognized in Middle English in the bipartite division into Early and Late Middle English, with the division being (variably) somewhere in the first half of the 14th century. Yet, even within these early and late periods, time-depth is an issue that ought to be considered for a proper diatopic interpretation of the linguistic evidence. In the period with which we are concerned, there was considerable linguistic change, notably in the phonology. Some of this is reflected in the orthography. But orthography itself might change too, without implication of phonic change, e.g. the deselection of the Old English letters ‘ƿ’ and ‘ð’. And change in orthographic practice may provide markers of regional differentiation. The most marked case is the lack of distinction to be found in Late ME northern texts between and , when such a distinction is to be found in the midland and the southern texts. (See Benskin 1982 for the palaeographical and orthographic history of this development; and see Map 31.6 THEY below for illustration.) Grammars and handbooks that treat substantially of Middle English (most significantly those of Luick [1921], Jordan [1968], Mosse´ [1968], the Cambridge History of the English Language, Vols. I and II [Hogg 1992; Blake 1989]) have recognized and described linguistic change through the period, particularly in the phonology: Trisyllabic Shortening, Open-Syllable Lengthening, Degemination of Consonants, Loss of final “schwa”, Northern Fronting, the early changes that are customarily bundled under the “Great Vowel Shift”. However, these diachronic developments are placed in relation to the linguistic geography in a broad way, if at all, e.g. as “Northern” (e.g. Northern Fronting) or “West Midland” (e.g. → [ɔ], as the reflex of Grmc. *a before a single nasal or a single nasal plus voiceless consonant). With the exception of The Cambridge History volumes, the major and most-referred-to accounts were written long before the detailed diatopic conspectus offered by LALME. Even within LALME, ostensibly covering the period 1350 to 1450, there is a time-depth which is not explicitly recognized in the data. Apart from the issue of the time-spans of both LALME and LAEME, there are sources used in LALME written before 1350 and after 1450. The maps in LALME are underlyingly diachronically skewed: data collected for the South generally belong to the beginning of the period covered by the atlas, while those for the North belong to the later part. Also, in the South the language of texts was affected earlier by the spreading standard; the bulk of texts containing northern Middle English survive from a later date. What we see on a LALME “dot map” (LALME 1, McIntosh et al. 1986: 305–551) or “item map” (LALME 2, McIntosh et al. 1986) looks like a synchronic presentation of data across geographical space, but in fact the maps contain substantial variation in the dates of the linguistic forms evidenced at the different survey points. In LAEME (Laing and Lass 2008), the geographical coverage provided by the sources is not uniform: extant texts for the North and for the South are scant, while they are quite numerous for the West and East Midlands, but they seem to be totally lacking for the Central Midlands.
31 Middle English: Dialects
4 Reconstructing the dialect continuum for Middle English When dealing with language from the late Middle Ages all our data come, of necessity, from the written language. Our witnesses for the past stages of the language were writers and copying scribes: we have to deal with “native writers”. Their written output is our only directly available source material. Our evidence comprises bundles of linguistic objects found in texts: Middle English is a “text language” (Fleischman 2000: 34) and a manuscript may contain one or more “text dialects”. Texts, therefore, take the place of the informants who can be questioned directly in a modern survey. The sample of witnesses is accidental – whatever texts have happened to have survived until the present. Unlike in a modern dialect survey, we cannot return to an informant to ask for more examples or for elucidation about a form. We have to infer anything about the spoken language of the writers and copiers of the texts from orthography or from linguistic patterns such as rhyme or alliteration. We also have to contend with different types of witness. The witnesses are not homogeneous in their general character. However, they may be divided into two basic types: (1) primary, being those which provide sufficient extralinguistic evidence to allow a prima facie localization; and (2) secondary, being those which offer valuable linguistic information, but whose provenance is not determinable from any extralinguistic evidence. Typically, primary witnesses are legal documents: charters, dealing with such matters as leases, bonds, and alliances, marriages, testaments, and also record books (ecclesiastical and lay) recording the proceedings of courts and local administrations. As a class, these texts may be termed “local documents” (Benskin 1977, 1981). They offer evidence of local associations through references in their text to persons and places, and they are usually dated. Secondary witnesses are generally literary texts. For the most part these do not offer information about their provenance, albeit there are exceptions (such as Dan Michel’s Ayenbyte of Inwyt in London, British Library, Arundel 57) which allow such texts to be used as primary witnesses. Literary texts are usually copies, which may or may not reflect the language of their copier, and their linguistic provenance therefore has to be discovered through a process of comparison with the language patterns afforded by the primary witnesses. For LALME and LAEME, this was done by a method called the “fit”-technique (LALME 1 McIntosh et al. 1986: 9–12; Benskin 1988, 1991). In the text to be localized is identified a set of linguistic features known to be diatopically salient from material already mapped from the primary witnesses or from already localized texts. By progressively eliminating areas of the map where these features do not occur, the provenance of the text may be discovered. One aims to have at the end of the process some single – preferably small – area which has not been eliminated, where only that set of features from the text being fitted occurs as an “assemblage”. Where each feature may have different, albeit overlapping, distributions, the aim of the procedure is to discover a locality where in their areas of distributions they all overlap together. The overlap defines the area of distribution of the assemblage of those features. Thus, any text not yet localized which contains that assemblage is likely to have been written in that area and/or by someone from that area or it is a faithful copy of a manuscript text written in that area and/or by someone from that area. How do we obtain linguistic information from the witnesses?
485
486
IV Middle English – Questionnaire Method (QM): we formulate a questionnaire (as for a modern linguistic survey) and record for the set of items the written forms which realize each item as they occur in each text. The result is a “linguistic profile” (LP) for each text. (See Table 31.1.) – Tagging Method (TM): we transcribe texts, key them to disk and, with the aid of a computer program, we tag each word and morpheme of the text lexicogrammatically. The QM is aimed at systematically collecting attestations of a predetermined subset of linguistic items whose forms are believed to show diatopic variation. In the TM, material is collected beyond that required for making dialect maps. But from the corpus created using the TM, diatopically salient material can be mined and because full running texts are recorded, each word and morpheme retains its linguistic context. With the QM, the collected forms are decontextualized. LALME (McIntosh et al. 1986) was made using the QM and LAEME (Laing and Lass 2008) and A Linguistic Atlas of Older Scots (LAOS, Williamson 2008) using the TM. Table 31.1: Exemplar linguistic data from three Linguistic Profiles in A Linguistic Atlas of Late Mediaeval English, 1350–1450, vol. 3 (McIntosh et al. 1986) LP 494 London, British Library, Add. MS 25013. Yorkshire, West Riding 399 428
LP 4629 London, British Library, Harley 2390. Norfolk 602 283
LP 7660 Leeds University, Brotherton Library 500. Worcestershire 370 266
SHE HER THEIR THEY THEM NOT
scho hir yair ( yaire ) (( yar )) yai yame yaim (( yaim )) noght ( not )
he (( a )) her ( here ) her here þey he ( heo þei ) hem not ( nout ) (( nouȝt ))
Present Participle FIRST
-ande -and ( -yng )
sche hyr here ( here ) here yeir ( here ) yei þei hem (( them )) nouȝth nootȝ notȝ nouȝt nowȝt nattȝ nowthȝ natȝ nat nottȝ nouwȝt ( natt ) -yng -yngg
AFTER IS ARE
after (( aftyr )) is es are ar ( er arne )
CHURCH EACH FIRE
kirke ilk ( ilke-a ilk-a iche-a ) (( ylke )) fire
FROM MAN
fro man (mon) (( mon ))
ITEMS
first (( fyrst frist ))
fyrstȝ fystȝ; ye-ffyrstȝ þe-firstȝ affttyr (( afftter )) is (( his )) ben ( arn ) (( harn ben )) chirche ich iche (( eche )) feyr feer fer ferr fyer ffer ( fyir ) from man (( man ))
-ynge furst ( furste fyrste ); þe-furste ( þe-firste ) aftur is beþ buþ buth ( but ben ) chyrche ( churche cherche ) vche vch ( vche-a eche ech ech- yche ich- ) fir fuyr ( fyr ) from ( form from fro ) mon (Continued )
31 Middle English: Dialects
487
Table 31.1: Continued LP 7660 Leeds University, Brotherton Library 500. Worcestershire 370 266
LP 494 London, British Library, Add. MS 25013. Yorkshire, West Riding 399 428
LP 4629 London, British Library, Harley 2390. Norfolk 602 283
MANY MUCH
mony (( many )) mykel ( mikil mykill mykil mikell mikel mykyl )
SHALL
sal (( sall schal )); sal (sall schal)
many manyȝe mony (( many monye )) mechel meche ( muche ) mechell meche mechyll (( mochell moche mecyll )) schal (( chal )); xall ( xal schall ); schul schulle (schele) xullyn xullyn (( schule schullen )) (xall xulle) xulde ( xullde ); schulde xulldyn xuldyn sweche swiche suche ( sych syche ) ( soche ) (( syge )) wen ( wenne ) qwan quan qwane (qwanne qwann qwanne) qwere qwere wer were wher qwere– qwethir qweyr qwethir weþer ( wheþer wer ) qweyre qwedere qwiche qwich þe-weche weche þe-wyche ye-qwiche qwyche wyche ( þe-whiche whiche þe-wheche wiche )
ITEMS
SG; PL
SHOULD SG; PL
SUCH WHEN
WHERE WHETHER WHICH
WORLD
schuld (( sulde suld )); schuld (( sulde )) swilk ( suche ) (( swilke )) when (( wen whan ))
whare ( whar whore where ) whether ( wheyer ) whilk ( ye-whilk ye-whylke whilke ye-wilke wilke ye-whiche wiche weche ) worlde ( worde ) (( word world ))
werld
worlde (( wrolde ))
[Note: Parentheses reflect the frequency of occurrence of the forms as recorded: unbracketed forms have the dominant frequency, forms in () occur “about one third to two thirds as frequently as the dominant form” and forms in (()) “occur about less than about one-third as commonly as the dominant form” (LALME 3, McIntosh et al. 1986: xiv).]
In Table 31.1 are listed the “forms” recorded in three manuscript texts for 28 “items” in the LALME questionnaire (reproduced here with the kind permission of the authors). (Note that in the table I have put the respective sets of singular and plural forms for SHALL and SHOULD in single rows.) In the discussion which follows I will use the following terms: “item”, “form”, and “feature”. An item is a test unit for linguistic comparison. Each question in a linguistic-survey questionnaire to which informants respond is an item. It is a superordinate for a group of forms, which are functionally equivalent. These can be grouped together as being: “the same word”, e.g. , , = SHE; or “the same morpheme”, e.g. , , , = 3P SG PRES IND “inflection”. In these cases, form is more or less equivalent to “the spelling adopted for the
488
IV Middle English word or morpheme”. But an item may also comprise a group of forms having “the same basic meaning”, e.g. til, fort, unto, þat = UNTIL. In this case, form is the equivalent of “the word chosen to express the meaning”. A linguistic feature is a segment comprising all or part of a form, or a set of forms, which realize one or more items: e.g. initial in forms of SHE; for the 3P SG PRES IND inflection; fort-forms expressing the meaning UNTIL. Thus, in Table 31.1, the questionnaire items are listed in column 1. In columns 2 through 4 are listed the forms corresponding to the items in column 1 which were found in texts in the three manuscripts whose designations are given in the column heads. Also given for each manuscript is the county in which the LP has been localized in LALME and the grid reference of its localization.
4.1 The methodology for LAEME LAEME (Laing and Lass 2008) has been constructed using the transcription-andtagging method. Complete manuscript texts or extensive samples from very long texts have been transcribed to disk. The transcribed texts are then lexico-grammatically tagged: each word and morpheme is assigned a tag comprising a “lexel” (lexical element) and a “grammel” (grammatical element), e.g. $thought/nOd_YOT
The start of the tag is marked by “$” and the end by the tie “_”, which links the tag to the manuscript word it describes, in this case “YOT”. The lexel for “YOT” is “thought”, which is a modern English gloss. The grammel (separated from the lexel by “/”) is “nOd”, where “n” = “noun” and “O” = “object” and “d” = “direct”. The grammel describes the grammatical category and function of the text-word at the point in the text at which it occurs, so “YOT” here is a noun functioning as a direct object. An example of a stretch of tagged text is: $/P11N+V_*HIC $be/vps11_AM $/AN_A $clerk/n_CLERC $/RTA_YAT $haunt/vps13F_HAUnT+ES $/vps13F_+ES $school/nOd{rh}_SCOLE “I am a cleric that haunts [frequents] [the] school”.
The texts are initially tagged as running text, but from this text, a dictionary version is made. In this “text dictionary”, the words and morphemes are arranged in alphanumerical order by tag and after each form is a frequency count of its occurrence when associated with its tag, e.g. $such/aj SUILC 1 $such/ajOd SUILC 1 $sweet/aj SUYT 1
31 Middle English: Dialects $sweet/aj-voc SUYTHE 1 $than/cj YAN 4 $that/cj YAT 6 $therefore/av-k YAR-FOR 1 $therein{p}/av-k{rh} YAR-InNE 1 $thing/nOd{rh} YInK 1 $to/im+C *TO 2 TO 2 $to/im+H *TO 1 $to/pr TO 2
The text dictionaries parallel the Linguistic Profiles in LALME, and from their data, features may be extracted for mapping. More information about the LAEME corpus and the principles of tagging can be found in the LAEME website (Laing and Lass 2008, Introduction, Chapters 3 and 4), and in Laing and Lass (2006). Examples of maps in LAEME can be found by accessing “Maps”. The LPs and text dictionaries are like the completed questionnaires in a modern survey. The items, as a set, are applied to all the witnesses to discover what forms attest each item. In this way, the items elicit a description of the linguistic variation. Because the same items are applied to each witness, they provide a framework for comparison of the materials. However, in a modern survey it is usually possible to obtain responses to all the items in the questionnaire. In reality, a medieval text will lack information for questionnaire items: if any questionnaire item is not attested by any form in a text, then we can have no response. The type of text dictates the measure of response: a long literary text will return a fuller questionnaire than a single-sheet charter or a will. And length of text is not the only factor here: the subject matter inevitably constrains the scope of the vocabulary. The inevitable lacunae across the corpus of LPs compromise the comparability of the data to some degree. That said, in evaluating the lacunae for diatopic purposes, it is better, nevertheless, to follow the maxim that “absence of evidence is not evidence of absence”.
5 Dialectal variation in Middle English 5.1 Regional differences How do we discover what is regionally marked in Middle English texts? As with a modern linguistic survey, we make maps. By displaying the occurrence of any linguistic feature on a map, we can see the range of its distribution – whether it is found over the whole area of survey, whether it is confined to a small area, or a much larger area, whether its distribution is dense and concentrated or sparse and diffuse, and whether a feature occurs in different areas, geographically separated. The maps reveal to us the varying diatopic saliences of linguistic features. In this section are presented maps which illustrate some key distributions and also different kinds and extents of distribution. The data in the maps are largely from LALME (McIntosh et al. 1986; reproduced with the kind permission of the authors). Materials from LAEME (Laing and Lass 2008) will also be referred to. What is presented is perforce highly selective. The maps have been drawn in such a way as
489
490
IV Middle English to provide a comparative presentation. The method used also schematizes the distributions somewhat and some comment about the method chosen is required. Much of the literature on the dialectology of Middle English and Modern English refers to the distribution of linguistic characteristics in terms of “isoglosses”. In discussions of the relationship between the two periods, the notion of “shift” is invoked (e.g. Fisiak 1983; Dietz 1989): for example, the area of distribution of the modern dialectal reflex(es) of a characteristic such as Middle English /aː/ and /ɔː/ < Old English a¯ has changed. An isogloss is a cartographic construct, intended to show the geographical boundary between the distributions of two linguistic features. But the isogloss is problematic: it often distorts what typically happens when the bounds of two distributions meet, by sanitizing or idealizing the co-occurrence of the bounds. There is rarely a clear-cut line between the distributions of features. Rather, they may often overlap, so that there are areas when both forms are used to a greater or lesser extent in neighboring communities or used by the members of one community and understood by those of another. Where the boundaries of features meet, we find zones of transition. It is these zones of transition, shifting across space, which define the dialect continuum. In LALME, the “dot maps” and in LAEME and LAOS, the “feature maps” display distributions as patterns of symbols at the locations where features have been attested, without attempting to demarcate the bounds of the distributions. What lies between the survey points is unknown. An isogloss makes assumptions about what is or is not there. To present information about features in the maps for this chapter, I have resorted to a variant of the isogloss – the “heterogloss”. A heterogloss is a line which attempts to demarcate the bounds of the distribution of a single feature. Where the bounds of the distributions of two features meet, there will be two boundary lines – one for each feature. Heteroglosses allow us to see where there are overlaps between the distributions. They still make assumptions, of course, about what is in effect terra incognita. (On the notions of “isogloss” and “heterogloss”, see Chambers and Trudgill 1998; and for an excellent critique of the notion of “isogloss”, see Kretzschmar 2003.) The maps offered here provide simplified pictures of the distributions of the features to be discussed. Nonetheless they capture some of the complexities of patterning. The reader is strongly encouraged to look at the original maps in LALME (McIntosh et al. 1986) and in LAEME (Laing and Lass 2008) on which the maps here presented are based. Certain geographical areas are more clearly “dialectal” than others and may admit more or less complex variation within. The North – geographically north and east of a line that could be drawn from North Lancashire to the Humber Estuary – seems clearly defined by a number of linguistic features, although there are variations in the boundaries of their distributions. For Late Middle English, although there is evidence of standard forms, these are less apparent than in the South and Midlands. Early Middle English evidence for the North is not abundant and is geographically sparse (see the keymap for LAEME, Laing and Lass 2008). For the Late Middle English period, more texts are available, especially in the 15th century, in which more documentary texts are available. One of the defining shibboleths for a northern text is whether it contains a-type spellings in reflexes of OE a¯. LALME collected spellings of these reflexes under an item A, O. Map 31.1 presents heteroglosses for a-type spellings and o-type spellings. (It should be noted that this item was collected only for the Northern part of the LALME survey,
31 Middle English: Dialects covering the North and Midlands.) The distribution indicates a well-defined southern boundary for a-type spellings. The heterogloss runs north-west to south-east from the Ribble, meandering through the West Riding and taking in North Lincolnshire. There are also two coherent islands of a-type forms in East Central Lincolnshire and in South Central Lincolnshire respectively. However, o-types are quite abundant north of the a-heterogloss. First, the northern bounds of the main o-distribution overlap the southern bounds of the a-distribution, so that there is a zone where o- and a-types cooccur. Further, within the area of the a-distribution are areas with o-also. The prevalence of o-type in the North within the a-area cannot be taken to reflect the spoken language. Rather, in most cases they are to be found there because the data are from written texts. The o-type may be prevalent because many of the northern texts are legal documents (hence subject more to contemporary and external linguistic influences, including
Map 31.1: OE a¯ reflexes spelled with either a or o, e.g. , , , or
491
492
IV Middle English features of standardization) and show some degree of use of non-northern forms (see the comments below on AT INF and TIL INF). While comparison of the medieval and modern boundaries of the a-type might be a feasible exercise, comparison of the oboundary must be more problematic. In any case, bundling reflexes of OE a¯ into a single etymological category and mapping the result is not a particularly enlightening exercise. Rather, one would want to map the distributions of each reflex individually and observe any variations in the patterns. Within the North, clear patterns of variation are hard to discern. Many of the northern texts used for LALME are from local documents and so the texts are brief and provide limited evidence with many lacunae for the LALME questionnaire items. The LALME maps for some items conceal some interesting differences between text types. For example, dot map 688 in LALME (McIntosh et al. 1986) shows the distribution of AT INF, a nice Northern feature (of Scandinavian origin). However, closer analysis of the kinds of texts in which this occurs, show AT INF to be confined almost exclusively to literary texts and to be almost absent from documents. TIL forms, collected under the LALME item TO INF, occur only sporadically in the local documents (see Williamson 2004: 267–272). This difference may be a product of the earlier date of many of the literary texts. Even if they are later copies, they may preserve earlier forms. Also there may have been a tendency amongst 15th-century writers of documents with legal training to eschew what might be considered “marked” or “provincial” forms in their professional writings. Lincolnshire evidences interesting patterns of distribution, sharing many features with the North, but not others. For example, among the salient Northern features, -and for the present participle is to be found across Lincolnshire, but not s(c)ho for SHE. (See Map 31.2 SHE and Map 31.3 Present Participle.) Indeed, some northern features may have receded from Lincolnshire in the post-medieval period. For example, Lincolnshire in the medieval period may have fallen within the bounds of the area of Northern Fronting (where OE, ON o¯ > /øː/ or /yː/, but the affected forms were later replaced from an area where this change did not take place; see Britton 2004). South Lincolnshire seems to be part of the nexus of dialectal complexity that is to be found around The Wash, running through Ely and into Norfolk and South Lincolnshire. Norfolk itself is dialectally complex as evidenced in the LALME materials. There are a number of distinctive local features in this area: qwh– type as reflex of OE hw, where further north qu-, qw- is to be found; spellings which imply /θ(t)/ for /xt/ as reflex of OE -ht, e.g. mith(t) ‘might’, nouth ‘nought’. A distinctive form to be found in Norfolk and in North Suffolk is for the initial consonant in SHALL and SHOULD, e.g. , . A distinctive marker of northern texts are the types swilk SUCH, whilk WHICH and ilk EACH. These have very similar bounds to their southern distributions and take in Lincolnshire. In the case of swilk and whilk, North Norfolk forms part of their distributions, while ilk also takes in south central Norfolk and an area of north central Suffolk. Within the period covered by LALME, the written language of London underwent an evolution, partly shaped in different ways at different periods by immigration from East Anglia and the East and Central Midlands. The stages of development are outlined in Samuels (1963). From London language, influenced by external and internal sources, emerged the standard which was to spread more widely across England, interacting
31 Middle English: Dialects
Map 31.2: SHE: scho-, sho-, he(e-, heo-, and hi-types
with local forms of language. Note that this should not be termed, as it often is, “Chancery Standard”, for the Chancery used predominantly Latin and such English that it dealt with came into it in texts from different regions of the country. Chancery copies of texts originating in different parts of the country often preserve their regional characteristics of language. The Chancery could not have been, nor was, the crucible for the formation of what became the standard. For a crucial corrective to the “Chancery Standard” fallacy, see Benskin (2004). As it spread across the country, the standard was adopted in varying measure, mixing with more local forms and diluting them. Many texts especially in the South in the 15th century show varying degrees of standardization in their language. A text’s local credentials might depend on forms which are in the minority as compared with the standard forms they might contain (cf. Schaefer, Chapter 33).
493
494
IV Middle English
Map 31.3: Present participle: types with -and, -end, and –ind
“Regional standards” or at least conventions could develop. The Central Midlands is dialectally not particularly rich, partly because of a tendency to leveling out of very local features in the extant texts. Many of the surviving texts from this area are Wycliffite and in their production there seems to have been an inclination to avoid highly local forms. In the Early Middle English period, no texts written in English are to be found for this area. The dialectology for this area for the period remains terra incognita. The West Midlands and East Anglia, by contrast seem to be the richest dialect areas for surviving texts. Texts survive in abundant quantity for the West Midlands throughout the Middle English period. Indeed, it seems to have been an important centre in the reinvention of written English. (See the Introduction to LAEME, Laing and Lass 2008, and also references therein.) The West Midlands shows a number of distinctive features, uch
31 Middle English: Dialects and euch for EACH, mony for MANY, fuir for FIRE, beoth for BE PRES PL. It was the focus of a great deal of interest in earlier studies, notably those of Oakden (1930) and Serjeantson (1927).
5.2 Exemplar maps One of the items which shows remarkable variation in its written forms in LALME is the item SHE. The most common type is s(c)he, where she is the standard form. The vowel aside, there is a major distinction between forms with sch- or sh- and h-. The h-type, of course, retained the Old English initial consonant. The s(c)h-type, with a palatal fricative, may well have developed from forms with initial h- (see Britton 1991). Map 31.2 shows the distributions of h-type with or or and of sch- and sh-types with vowel . The types with palatalized initial consonants are Northern. The sch- and sh-heteroglosses show a similar limit for their distributions with the sh-distribution having a more westward boundary into East Cheshire and slight eastward extension into Lincolnshire. The heo and he(e types are found primarily in the West Midlands and South-west, also extending into South Hampshire and into South-west Sussex. Their distributions show similar limits in the Southern part of their eastern boundaries, but heo extends further north, covering Shropshire. There is a small pocket of he(e type in South-East Suffolk. hi, hy are found in North Kent. A long-recognized division in Middle English is between forms of MANY with o and forms with a for the tonic vowel. Map 31.4 displays heteroglosses for these two types. However, the division is not clear-cut and there is some overlap between the two distributions on the western boundary of many (broken line) and the eastern boundary of mony (solid black line). Indeed, in the South-west Midlands, the many distribution extends as far west as the Welsh Marches, while the mony boundary extends eastward into North Wiltshire. This wide area of overlap between the two types seems to have been present also in the Early Middle English period (LAEME map “unpublished”; but in the revised version users can create a map for this feature). There is also an extensive “island” of the mony type further east, in South-central England. The northern boundary of the many type extends as far North as Westmorland in the North-west and south-eastwards through the North Riding and into the East Riding of Yorkshire; mony forms are also found in the North-east of England – Eastern Cumberland and Durham and Northumberland. In late 14th-century and 15th-century Scots mony is to be found throughout. This North-eastern distribution may well be connected historically to the Scots one. In the LAEME data, there is no evidence of mony type in the North-east, but this may be due to lack of evidence for the area surviving from the earlier period. In the LALME data, there is a zone between the northern boundary of the many distribution and the southern and western boundary of the mony distribution in the North-east for which no forms for MANY are attested in the sources localized there. This is represented on the map by “?”. The mony / many opposition is cartographically rather complex. A third type for MANY is meny. This occurs in the South-west, although its distribution extends into Herefordshire and overlaps with the many and mony type distributions there. It extends at its eastmost into the
495
496
IV Middle English
Map 31.4: MANY: forms with a, o, or e for the tonic vowel
south-west corner of Sussex. Islands of meny are to be found further east, including south-west Essex and south-west Suffolk. The meny type is not evidenced at all in the LAEME data. The personal plural pronouns THEY, THEM, and THEIR are of considerable dialectal interest. Two basic types for each one may be identified: (1) those of Old English origin, which we can characterize from their initial consonant, /h/; (2) those of Scandinavian origin, the ancestors of the standard forms, with an initial /ð/, with the principal spellings þ, y, or th (others are possible, especially in Early Middle English). The dialectal interest arises, first, from the relative distributions of the types. The second from the spelling of the /ð/ segment in the forms of Scandinavian origin. By the Late ME period the /h/ type was being replaced by the /ð/ type. The Early Middle English evidence shows little penetration southward by the latter type (LAEME), but by the ME period
31 Middle English: Dialects (as evidenced in LALME), the /ð/ type shows evidence of penetration into the area of the original /h/ type. With THEM and THEIR, the progress is rather diffuse in the South. However, THEY with /ð/ appears widely across the area and overlaps with the /h/ type. This is represented in Map 31.6, where the distribution of /ð/ type with spellings in þ, y is contrasted with the distribution of the /h/ type. (th spellings are not included and are widely scattered anyway.) The northern boundaries of the /h/ distributions of THEM and THEIR are shown in Map 31.5. It is clear that THEY /h/ type has receded much further south and westwards and the /ð/ type for THEY is much more progressive. The distribution of THEY also allows comment on a case of dialectal distribution of written forms: y as against þ. In the representation of /ð/ and /θ/, as well as th, y and þ were used. Where th is to be found widely across the country, there is a geographical division between the use of y and þ. This is evident in Map 31.6 where forms with y are to be found in the north extending southward to Lincolnshire, Nottinghamshire, and North Derbyshire. y is also to be found in Norfolk. þ is found South of this, but also its distribution overlaps that of y in the southern half of Lincolnshire and in Nottinghamshire and North Derbyshire. The use of y for /ð/ and also for /θ/ arose from developments in scripts in the North and North Midlands. In the formation of the littera þ, so that its figurae became more y-like and came eventually to be identified with those of y. In Older Scots also, y is used commonly for /ð/ and /θ/. y continued to be used to represent high front vowels, of course, even in Northern Middle English and Scots. Middle English exhibited dialectal variation in its morphology. A salient case is that of the present participle as shown in Map 31.3. The standard -ing type is to be found geographically widespread in Middle English texts. Of some diatopic significance is the -and type, found in the North, Lincolnshire, and over most of Norfolk. There are islands of this type in the North-west Midlands. It is to be found also in Devon, where there seem to have been scribes at work whose language contains some elements which are clearly Northern in origin. Other types are -ind and -end. All these types are descendents from Old English. The -end type is found predominantly in Norfolk, Ely, and Cambridgeshire. There is also an island in Essex and another in North-west Gloucestershire, falling within the main area of the -ind type, which is a South-west Midland characteristic. All the types, -and, -end, and -ind, are to be found in the London area, with the -ind type widest spread, extending into North-west Kent. The verb BE exhibits dialectal variation in many of its forms. Map 31.7 displays the distributions of variants for forms for the present plural, corresponding to Standard English are. The are type, corresponding to the standard form is widespread. Of dialectal interest is the er type (from Old Scandinavian). This is found in an area delimited in the South by a border running from northern Lancashire, meandering through the West Riding and looping to take in North and East Nottinghamshire and south and east around the borders of Lincolnshire, so that all of Lincolnshire falls within the er distribution. er extends north to take in most of Cumberland and Westmorland and eastern Durham, but western Durham, eastern Cumberland, and most of Northumberland fall outside, having the ar type, which is shared with Scots. The ben type is found across all of the South and Midlands. Its northern boundary partly intertwines with that of the southern boundary of the er type, creating areas of overlap or transition between the two types. The northern boundary of
497
498
IV Middle English
Map 31.5: THEM, THEIR: Northern limits of h-types
ben cuts through the centre of Lincolnshire, so that south Lincolnshire falls within both the distributions ben and er. The beth type is found rather further south. Its distribution embraces all of the West Midlands, taking in the eastern half of Shropshire and all of Staffordshire and Warwickshire. In the east, it takes in Essex and Suffolk and part of South Norfolk. A fourth type, beoth, is much more local, covering an area of the West Midlands, parts of Herefordshire, Worcestershire, and North Gloucestershire. The beon type (not mapped) is also characteristic of the West Midlands, but encompassing a smaller area, mainly Herefordshire and Worcestershire. The buth type also occurs in the West Midlands, but has a wider distribution into the South-west and the South, taking in South Wiltshire, Hampshire, and Sussex.
31 Middle English: Dialects
Map 31.6: THEY: Northern limits of þ and h and southern limit of y
The present tense indicative endings of Middle English have broad regional distributions: in the South and West Midlands, 3P SG/PL -th type; in the Central and West Midlands (including East Anglia) 3P SG -th, PL -en or -e; in the North -is ~ ∅ [i.e. no inflection]. The Northern present tense indicative system varies the ending according to the category and adjacency to the verb of the subject. If the subject is not a personal pronoun or if the subject is a personal pronoun but not adjacent to the verb, then the verb is inflected, usually with the -is type inflection (but other forms are possible, -us, -es). If the subject is a personal pronoun and it is adjacent to the verb, then the 1P and 2P SG and 1P, 2P, 3P PL are uninflected, but the 3P SG has -is. However, an area, covering South Lincolnshire, Rutland, and Leicestershire, contains a paradigm that uses formally Midland-type inflections, but patterned similarly
499
500
IV Middle English
Map 31.7: The present plural of BE: er, ben, beth, and beoth types
to the Northern Present Tense Rule: if the subject is not a personal pronoun or is a nonadjacent personal pronoun, the 3P SG and 1P, 2P, 3P PL have -eth (i.e. the singular inflection appears to be extended to the plural); if the subject is a personal pronoun adjacent to the verb, then the 3P SG has -eth, the plural has the regular Midland -en or -e. For a full account, see McIntosh (1983). Map 31.8 reflects the history of OE y¯. It shows the distribution of types of spelling for the LALME item FIRE. The ui-type, found mainly in the West Midlands would seem to reflect the retention of a rounded vowel, although, of course, spellings can remain after phonological change. There are also islands of this type across the South, including south-west Essex. The e- and ei-types reflect the earlier (9th century) change of unrounding, presumably of a mid or mid-high front rounded vowel, /øː/ or /ʏː/, in East Anglia and in Kent. Elsewhere, OE y¯ later unrounded as a high front vowel, /iː/, reflected
31 Middle English: Dialects in such spellings as , . The ei- type may partly be a variant of the e-type or the i-type with an epenthetic vowel before /r/. The i-type is general across the North. It is hoped that the above examples will have given some idea of the patterns of diversity and richness in Middle English dialects.
6 Dialects and word geography While we have gained much knowledge of geographical variation in orthography and morphology, vocabulary has proved less tractable. Until recently the most extensive study was Rolf Kaiser’s (1937) Zur Geographie des mittelenglischen Wortschatzes.
Map 31.8: FIRE: ui-, e-, ei-, and ie-types (representing reflexes of OE y¯)
501
502
IV Middle English Angus McIntosh had also discussed issues and strategies in dealing with the lexis (McIntosh 1973, 1977). The dialectology of the lexis seemed to pose quite different problems from those of the other levels of analysis: orthography, phonology, and morphology. However, a major study by Richard Dance (2003) has offered a valuable insight into the vocabulary of Scandinavian origin and the interpretation of its dissemination. A more extensively-oriented project for the study of Middle English geography is being carried out by Marı´a Jose´ Carrillo-Linares and Edurne Garrido-Anes (2005–06, 2007, 2008, 2009) at the University of Huelva, Spain. This has already advanced understanding of, and offered solutions to, specific problems of establishing the geography of the vocabulary. The method of Carrillo-Linares and Garrido-Anes is to construct a database in which are recorded the occurrences of a set of words in the different manuscript copies of the same work. Comparison is made to see if a word has been substituted by another at parallel positions in the text or if the text was altered in some other way (e.g. omission, paraphrase) that implies avoidance. But retention need not mean that a word might be dialectally acceptable to a copyist: a word might be substituted in some positions in the manuscript text yet retained in others, especially where it occurs in rhyming or alliterative position and the copyist chose not to seek or could not find an appropriate substitute that would preserve the phonic function of the word. Close examination of the texts and the patterns of distribution of the words is necessary. While the project is still in its early stages, it has already revealed interesting patterns, although the work also shows the treatment of word geography requires special strategies to uncover the diatopic patterning in a reliable way.
7 Further prospects As well as the on-going work on LAEME and the word-geographical research at the University of Huelva, nearing completion (autumn 2011) is a revision of LALME. The result will be a freely-accessible on-line version of the atlas (eLALME) as a companion web-site to both LAEME and the LAOS. The association of LAEME and eLALME will offer a powerful diachronic as well as diatopic conspectus to the study of Middle English. Of course, none of these works is in any sense definitive. Rather they are but the groundwork for further investigation into the complexities of regional variants and the processes of change and their dissemination across space in the Middle English period.
8 References Benskin, Michael. 1977. Local archives and Middle English dialects. Journal of the Society of Archivists 5: 500–514. Benskin, Michael. 1981. The Middle English dialect atlas. In: Benskin and Samuels (eds.), xxvii–xli. Benskin, Michael. 1982. The letters and in later Middle English and some related matters. Journal of the Society of Archivists 7: 13–30. Benskin, Michael. 1988. The numerical classification of dialect maps. In: Pieter van Reenen and Karen van Reenen-Stein (eds.), Distributions Spatiales et Temporelles, Constellations de Manuscrits, 13–38. Amsterdam/Philadelphia: John Benjamins. Benskin, Michael. 1991. The “fit”-technique explained. In: Felicity Riddy (ed.), Regionalism in Late Medieval Manuscripts and Texts, 9–26. Cambridge: D. Brewer.
31 Middle English: Dialects Benskin, Michael. 1994. Descriptions of dialect and areal distributions. In Margaret Laing and Keith Williamson (eds.), Speaking in Our Tongues: Medieval Dialectology and Related Disciplines, 169–184. Cambridge: D. S. Brewer. Benskin, Michael. 2004. Chancery Standard. In: Christian J. Kay, Carole Hough, and Irene´ Wotherspoon (eds.), New Perspectives on English Historical Linguistics: Selected papers from 12 ICEHL, Glasgow, 21–26 August 2002. Vol. II: Lexis and Transmission, 1–40. Amsterdam/ Philadelphia: John Benjamins. Benskin, Michael and Margaret Laing. 1981. Translations and Mischsprachen. In: Benskin and Samuels (eds.), 55–106. Benskin, Michael and M. L. Samuels (eds.). 1981. So meny people longages and tonges: Philological Essays in Scots and Mediaeval English presented to Angus McIntosh. Edinburgh: The Editors. Blake, Norman. 1989. The Cambridge History of the English Language. Vol. II. 1066–1476. Cambridge: Cambridge University Press. Britton, Derek. 1991. On ME she/sho: A Scots solution to an English problem. Nowele [NorthWestern European Language Evolution] 17: 3–51. Britton, Derek. 2004. Northern Fronting and the North Lincolnshire reflexes of ME /uː/ and /oː/. Language Sciences 24: 221–229. Carrillo-Linares, Marı´a Jose´. 2005–06. Lexical dialectal items in Cursor Mundi: Contexts of occurrence and geographical distribution. Selim Journal 13: 151–178. Carrillo-Linares, Marı´a Jose´ and Edurne Garrido-Anes. 2007. Middle English lexical distributions: Two instances from the Lay Folks’ Catechism. In: Gabriella Mazzon (ed.), Studies in Middle English Forms and Meanings, 85–100. Frankfurt: Peter Lang. Carrillo-Linares, Marı´a Jose´ and Edurne Garrido-Anes. 2008. Middle English word geography: Methodology and applications illustrated. In: Marina Dossena, Richard Dury, and Maurizio Gotti (eds.), English Historical Linguistics 2006. Vol. III: Geo-Historical Variation in English, 67–89. Amsterdam/Philadelphia: John Benjamins. Carrillo-Linares, Marı´a Jose´ and Edurne Garrido-Anes. 2009. Middle English word geography: External sources for investigating the field. In: Marina Dossena and Roger Lass (eds.), Studies in English and European Dialectology. Frankfurt am Main: Peter Lang. Chambers, J. K. and Peter Trudgill. 1998. Dialectology. Cambridge: Cambridge University Press. Dance, Richard. 2003. Words derived from Old Norse in Early Middle English: Studies in the Vocabulary of the South-West Midland Texts. Tempe: Arizona Center for Medieval and Renaissance Studies. Dietz, Klaus. 1989. Die historische Schichtung phonologischer Isoglossen in den englischen Dialekten. In: Andreas Fischer (ed.), The History and Dialects of English, 295–329. Heidelberg: Carl Winter. Ellis, Alexander J. 1869–89. On Early English Pronunciation. 5 vols. London: Asher. Fisiak, Jacek. 1983. English dialects in the fifteenth century: Some observations concerning the shift of isoglosses. Folia Linguistica Historica 4(2): 195–217. Fleischman, Suzanne. 2000. Methodologies and ideologies in historical linguistics: On working with older languages. In: Susan C. Herring, Pieter van Reneen, and Lene Schøsler (eds.), Textual Parameters in Older Languages, 33–58. Amsterdam/Philadelphia: John Benjamins. Hogg, Richard M. 1992. The Cambridge History of the English Language. Vol. I. The Beginnings to 1066. Cambridge: Cambridge University Press. Jordan, Richard. 1968. Handbuch der mittelenglischen Grammatik: Lautlehre. Heidelberg: C. Winter. Kaiser, Rolf. 1937. Zur Geographie des mittelenglischen Wortschatzes. Palaestra 205. Leipzig: Mayer & Mu¨ller. Kitson, Peter R. 1993. Geographical variation in Old English prepositions and the location of Ælfric’s and other literary texts. English Studies 74(1): 1–50.
503
504
IV Middle English Kitson, Peter R. 1995. The nature of Old English dialect distributions, mainly as exhibited in charter boundaries. In: Jacek Fisiak (ed.), Medieval Dialectology, 43–135. Berlin/New York: Mouton de Gruyter. Kitson, Peter R. 1997. When did Middle English begin? In: Jacek Fisiak (ed.) Studies in Middle English Linguistics, 221–269. Berlin/ New York: Mouton de Gruyter. Kretzschmar, William A., Jr. 2003. Mapping Southern English. American Speech 78 (2): 130–149. Kristensson, Gillis. 1967. A Survey of Middle English Dialects 1290–1350: The Six Northern Counties and Lincolnshire. Lund: University of Lund Press. Kristensson, Gillis. 1987. A Survey of Middle English Dialects 1290–1350: The West Midland Counties. Lund: University of Lund Press. Kristensson, Gillis. 1995. A Survey of Middle English Dialects 1290–1350: The East Midland Counties. Lund: University of Lund Press. Kristensson, Gillis. 2001a. A Survey of Middle English Dialects 1290–1350: The Southern Counties. I. Vowels (except Diphthongs). Lund: University of Lund Press. Kristensson, Gillis. 2001b. A Survey of Middle English Dialects 1290–1350: The Southern Counties. II. Diphthongs and Consonants. Lund: University of Lund Press. Laing, Margaret (ed.). 1989. Middle English Dialectology: Essays on some Principles and Problems. Aberdeen: Aberdeen University Press. Laing, Margaret. 2004. Multidimensionality: Time, space and stratigraphy in historical dialectology. In: Marina Dossena and Roger Lass (eds.), Methods and Data in English Historical Dialectology, 49–96. Bern: Peter Lang. Laing, Margaret and Roger Lass. 2006. Early Middle English dialectology: Problems and prospects. In: Ans van Kemenade and Bettlou Los (eds.), The Handbook of the History of English, 417–451. Oxford: Blackwell. Laing, Margaret and Roger Lass. 2008. A Linguistic Atlas of Early Middle English, 1150–1325 (LAEME). Edinburgh: University of Edinburgh. http://www.lel.ed.ac.uk/ihd/laeme1/laeme1.html Laing, Margaret and Keith Williamson. 2004. The archaeology of Middle English texts. In: Christian J. Kay and Jeremy J. Smith (eds.), Categorization in the History of English, 85– 145. Amsterdam /Philadelphia: John Benjamins. Lass, Roger. 2000. Language periodization and the concept “middle”. In: Taavitsainen et al. (eds.), 7–41. Luick, Karl. 1921. Historische Grammatik der englischen Sprache. Vol. 1. Leipzig: Tauchnitz. McClure, Peter. 1973. Lay subsidy rolls and dialect phonology. In: Folke Sandgren (ed.), Otium et Negotium. Studies in Onomatology and Library Science presented to Olof von Feilitzen. Stockholm: P. A. Norstedt & So¨ner. McIntosh, Angus. 1973. Word geography in the lexicography of medieval English. Annals of the New York Academy of Sciences 211: 55–66. [repr. in Laing, 1989, 86–97.] McIntosh, Angus. 1977. Middle English word geography: Its potential role in the study of the long-term impact of the Scandinavian settlements upon English. In: Thorsten Anderson and Karle Inge Sandred (eds.), The Vikings, Proceedings of a Symposium of the Faculty of Arts of Uppsala University, June 6–9, 1977. Uppsala: University of Uppsala. [repr. in Laing, 1989, 98–105.] McIntosh, Angus. 1983. Present indicative plural forms in the later Middle English of the North Midlands. In: Douglas Gray and E. G. Stanley (eds.), Middle English Studies presented to Norman Davis. Oxford: Oxford University Press, 235–244. [repr. in Laing, 1989, 116–122.] McIntosh, Angus, M. L. Samuels, and Michael Benskin. 1986. A Linguistic Atlas of Late Mediaeval English, 1350–1450 (LALME). Aberdeen: Aberdeen University Press / Edinburgh: Mercat Press. Moore, Samuel, Sanford B. Meech, and Harold Whitehall. 1935. Middle English dialect characteristics and dialect boundaries. Essays and Studies in English and Comparative Literature. University of Michigan Publications in Language and Literature 13: 1–60.
32 Middle English: Language contact
505
Mosse´, Fernand. 1968. A Handbook of Middle English. James A. Walker (trans.). Baltimore: Johns Hopkins Press. Oakden, J. P. 1930. Alliterative Poetry in Middle English: The Dialectal and Metrical Survey. Manchester: Manchester University Press. Orton, Harold et al. 1962–71. Survey of English Dialects. The Basic Material. 4 vols. Leeds: University of Leeds/E.J. Arnold. Samuels, M. L. 1963. Some applications of Middle English dialectology. English Studies 44: 81–94. [repr. in Laing, 1989, 64–80.] Serjeantson, M. S. 1927. The dialects of the West Midlands. Review of English Studies 3: 54–67. Taavitsainen, Irma, Terttu Nevalainen, Pa¨ivi Pahta, and Matti Rissanen (eds.). 2000. Placing Middle English in Context. Berlin/New York: Mouton de Gruyter. Williamson, Keith. 2002. The dialectology of “English” north of the Humber, c.1380–1500. In: Teresa Fanego, Bele´n Me´ndez-Naya, and Elena Seoane (eds.), Sounds, Words, Texts and Change. Selected Papers from 11 ICEHL, Santiago de Compostela, 7–11 September 2000, 253–286. Amsterdam/Philadelphia: John Benjamins. Williamson, Keith. 2008. A Linguistic Atlas of Older Scots, Phase 1: 1380-1500 (LAOS). http:// www.lel.ed.ac.uk/ihd/laos1/laos1.html Wright, Joseph. 1898–1905. The English Dialect Dictionary. 6 vols. London: H. Frowde.
Keith Williamson, Edinburgh (UK)
32 Middle English: Language contact 1 2 3 4 5 6 7
Introduction Language contact and change Multilingualism in the Middle English period Specific research questions Linguistic levels Conclusion References
Abstract The complex linguistic situation in the ME period, with widespread multilingualism and initial diglossia, has led to frequent contact-induced changes on all linguistic levels of English. The present chapter starts with a brief discussion of the relation between language contact and change and of the changing nature of ME multilingualism; then some specific research questions are introduced, such as the controversial issue of ME creolization, the frequent use of code-switching in medieval texts, and the possible Celtic influence on English. The remaining sections deal in some detail with contact-induced change on the various linguistic levels: while foreign lexical influence is well established, contactinduced structural changes are more controversial, since here a native origin is often equally possible. In many cases, especially of syntactic change, a polygenetic origin seems more plausible than a monocausal explanation. In any case, the extensive restructuring of Middle English cannot be explained without close linguistic contact. Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 505–519
506
IV Middle English
1 Introduction Speech communities hardly ever exist in isolation but have multiple contacts with other speech communities, with speakers of other languages and dialects, or with a culturally important “book” language. English has experienced a large number of such contacts with different ethnic and linguistic groups thoughout its history. From the beginning of their settlement, the Anglo-Saxons had contacts with speakers of Celtic languages, followed by contacts with Latin as the mainly written language of Christianity and culture, and with the spoken Scandinavian languages (here referred to as Old Norse). Since the linguistic results of the early contacts with Celtic and Old Norse only surface in ME texts, they are included in this discussion of ME language contact. For the ME period, intensive contact with French is one of the most obvious features, though Latin also extended its role into new domains (see Townend 2006). Additionally, there was trade-related contact with Low German and Low Dutch. However, not all ME dialects nor their speakers experienced language contact to the same extent, and we also find extensive interdialectal contact, resulting e.g. in the spread of Northern and East Midland features into Southern dialects. Even a superficial comparison of Old and Middle English reveals dramatic linguistic changes: Old English was an inflected and largely synthetic language with a predominantly Germanic lexicon. By the end of the ME period, English had lost much of its inflectional morphology, had become predominantly analytic, had undergone numerous syntactic changes, and had substantially enlarged its vocabulary. Contact-induced change played an important, though partly controversial, role in this extensive restructuring of Middle English.
2 Language contact and change Contact-induced change in general presupposes some degree of bilingualism or, in the case of dialects or closely related languages, mutual intelligibility. In situations of no or limited literacy, language contact predominantly happens in oral communication, as in the case of Celtic or Old Norse, and partly with French. Contact with Latin, on the other hand, primarily involved a written language of culture. The extent to which language contact triggers or reinforces change strongly depends on extralinguistic factors like the social prestige of and speakers’ attitudes toward a language, the intensity of contact, and the number of bilinguals. Foreign linguistic items may enter a language through borrowing or they may result from interference, often linked to imperfect learning. Contact-induced changes such as the attrition of inflectional morphology, however, do not necessarily lead to the importation of foreign features (for details, see Thomason 2001: Chapter 4). There is wide agreement that every linguistic feature can be borrowed under specific circumstances, though the probability of borrowing depends on various factors, an assumption which has led to the establishment of borrowing scales claiming, e.g., that non-basic vocabulary is borrowed before basic vocabulary, superficial phonological features, and grammatical features (Thomason 2001: 69–71). However, social and pragmatic factors, such as attitudes or imperfect learning, may overrule possible linguistic factors. Under specific circumstances, intensive language contact may result in a specific form of mixed language, such as a pidgin.
32 Middle English: Language contact Another contact phenomenon is shift to another language, which frequently entails language death. This led to the disappearance of Old Norse as well as to the adoption of English by the socially dominant, but numerically inferior elite of French speakers. Both languages, however, have left their traces in the English language. (See also Hickey 2010, Part II.)
3 Multilingualism in the Middle English period Throughout the ME period, England was a multilingual country, though individual bior multilingualism was never a general phenomenon and was very much socially and geographically conditioned. The geographical and social distribution of languages as well as their functions and status fundamentally changed over time. In the early part of the period we find a diglossic situation, in which different languages were predominantly used in specific domains. A description of this highly complex and changing situation has evidently to remain sketchy. Latin and French were the prestigious High languages during most of the period; the former was particularly used in religion, scholarship, education, literature, and administration, while the latter started as the language of the politically and socially dominant group and soon extended into domains such as literature, administration, and law, but lost its position in most of these towards the end of the period (see Kibbee 1991). English, on the other hand, began as the Low vernacular mainly spoken by the illiterate majority, but increasingly extended its functions, becoming the dominant language in most domains by late Middle English. The Celtic languages mainly survived in the fringes of Britain (see, however, Section 4.3), while the speakers of Old Norse had largely shifted to English by the late 11th century. In the first century after the conquest, Latin dominated as the written language of religion, scholarship, and literature as well as of the majority of official and legal documents, while texts in the two vernaculars are still relatively rare. French was the first language of a powerful minority, whose highest ranks were largely monolingual, while lower first-language speakers of French soon shifted to English, most likely after a period of bilingualism. A number of first-language speakers of English acquired French as a second language, but the majority, especially illiterate farmers and craftsmen, only spoke local varieties of English. For political reasons new groups of continental French speakers came to England around the middle of the 12th and the early 13th centuries. At this time the descendants of the Anglo-Norman conquerors were increasingly shifting to English; in the ensuing political tensions, language issues began to play a role, and some knowledge of English was more and more regarded as a proper mark of being English. The loss of Normandy (1204) furthermore severed the close ties of the nobility with France and affected the status of French, which had increasingly become a taught, but socially still highly prestigious second language. The functions of the written languages became more diverse from the middle of the 12th century on. In literature, French was preferred over English till the middle of the 13th century, mainly due to its role as the European language of culture; here as in most other domains English rapidly gained ground from the early 14th century onwards, a development furthered by the Hundred Years War. Latin dominates in religion and scholarship, and partly in administration, where English and French start to be increasingly used as well, while French is established as the language of the legal system.
507
508
IV Middle English The extent of English-French bilingualism in the 13th century is a matter of controversy, though by now English was widely known among the French-speaking gentry and many educated English speakers seem to have known some French. By the end of the 13th century, English had become the first language even for the vast majority of English-French bilinguals (for an earlier dating, see Townend 2006: 67). Though the general decrease of French is mirrored in a number of contemporary parliamentary decrees, French and Latin continued to be used next to the now-dominating English by professional scribes in various fields of administrative written discourse, such as official letters, petitions, and the Rolls of Parliament, into the first half of the 15th century. Thus, language shift to English in administration was a slow process, with a long period of multilingual usage upheld by a class of professional multilingual clerks. The final victory of English was also furthered by demographic and social changes: the growing urbanization up to the middle of the 14th century produced a wealthy Englishspeaking merchant class, while epidemics of the plague reduced England’s population by at least one third around the middle of the 14th century, leading to a shortage of labor and thus indirectly increasing the social prestige of English.
4 Specific research questions 4.1 Middle English – a creole? The contact-induced changes of Middle English are in general attributed to intensive borrowing but there have also been claims that extreme language mixing led to a type of new language (cf. Thomason 2001: 10): in such a view, Middle English (or some of its varieties) has variously been called a mixed language, a creole or a koine´. The controversial “creole hypothesis”, first advanced in the late 1970s (e.g. Bailey and Maroldt 1977), claimed that a creolization of English resulted either from the intimate contact between English and French in the south, or from that with Old Norse in the Danelaw. This is widely rejected now (see Trotter, Volume 2, Middle English Creolization), both because of the different sociohistorical situations in the ME period and in “classical” creolization contexts, and of the different linguistic outcome in Middle English: in spite of far-reaching regularization and simplification of ME grammar, there is neither complete loss of inflections, nor widespread loss of grammatical categories; Middle English even extended and reinforced certain grammatical categories (cf. Go¨rlach 1986). To explain the specific contact-situation in the Danelaw, Thomason and Kaufman (1988: Section 9.8) developed a model of “Norsification” of the northern Midland dialects of English, claiming that a fine-tuned concept of borrowing sufficiently explains this contact situation. Alternative explanations propose the development of an “Ausgleichssprache” between English and Old Norse, with neutralization of grammatical categories which differed in the two languages. Dawson (2003) interprets this process as koineization of two closely related, most likely mutually intelligible languages in the context of intensive direct speaker interaction.
4.2 Bilingualism and code-switching A large number of texts from the ME period are multilingual and mix two or three of the main languages, often within the same sentence. Such mixed-language texts
32 Middle English: Language contact neither reflect imperfect competence nor “corrupt” language, as claimed in earlier times, but are best accounted for as instances of written code-switching, a phenomenon well-known from many multilingual societies. Since code-switching is regarded as an important mechanism for contact-induced change (Thomason 2001: 129–136), these mixed texts not only illustrate the interaction between different languages in performance, but also provide important evidence for the working of ME linguistic contact. Code-switching occurs in a variety of ME text types and genres, such as sermons, scientific and medical texts, business accounts, legal texts, and letters; furthermore in literary texts, such as “macaronic” poems, drama, and travel accounts. In most cases, codeswitching either fulfils specific textual functions (e.g. emphasis, quotation, reiteration) or acts as a specific mode of discourse, as in example (1) below. The multilingual authors and scribes that wrote and copied such texts were certainly using these different languages on a regular basis, and their notion of a distinct or foreign language may have differed from our modern one (see Schendl 2002a; for a specific type of language mixing in medieval business writing see Wright 2011). The following samples illustrate switching into English in two non-literary text types, a Latin sermon and a French letter. Text (1) shows extensive intrasentential switching, a common feature in late ME sermons (see Halmari and Regetz 2011). (1)
Sermon: De celo querebant (early 15th century; Wenzel 1994: 274–275): Domini gouernouris most eciam be merciful in punchyng. Oportet ipsos attendere quod of stakis and stodis qui deberent stare in ista vinea quedam sunt smoþe and lightlich wul boo, quedam sunt so stif and so ful of warris quod homo schal to-cleue hom cicius quam planare. Quidam subditi sunt humiles and buxum, et de facili volunt corigi; […] ‘The lord’s governors must also be merciful in punishing. They should take notice that of the stakes and supports that should stand in this vineyard, some are smooth and will easily bend, others are so stiff and so full of obstinacy that a man will split them sooner than straighten them out. Some subjects are humble and obedient and will be easily corrected; […]’
Even more frequent is intersentential switching as illustrated in the French letter to Henry IV under (2). Here the switches into English carry a more personal and urgent note and seem to reflect the “we” code as opposed to the “they” code of the High language, French (see Schendl 2002b). (2)
Letter: Richard Kingeston to Henry IV (1403): Qar, mon tresredoute Seigneur, vous trouverez pour certein que si vous ne venez en vostre propre persone pour attendre [apres] voz rebelx en Galys, vous ne trouverez un gentil que veot attendre deinz vostre dit Countee. War fore, for Goddesake, thinketh on ȝour beste Frende, God, and thanke Hym as He hath deserved to ȝowe; and leveth nought that ȝe ne come for no man that may counsaille ȝowe the contrarie; for, by the trouthe that I schal be to ȝowe ȝet, this day the Walshmen supposen and trusten that ȝe schulle nought come there, and there fore, for Goddeslove, make them fals men. […]
509
510
IV Middle English Tresexcellent, trespuissant, et tresredoute Seignour, autrement say a present nieez. Jeo prie a la Benoit Trinite que vous ottroie bone vie ove tresentier sauntee a treslonge durre, and sende ȝowe sone to ows in help and prosperitee; […] ‘For, my most dread Lord, you will find for certain that, if you do not come in your own person to await your rebels in Wales, you will not find a single gentleman that will stop in your said county. Wherefore, for God’s sake, think on your best friend, God, and thank Him, as He hath deserved of you; and leave nought that you do not come for no man that may counsel you the contrary: for, by the truth that I shall be to you yet, this day the Welshmen suppose and trust that you shall not come there, and therefore, for God’s love, make them false men. […] Most excellent, most mighty, and most dread Lord, I know nothing besides at present. I pray the Blessed Trinity to give you good life, with most complete good health, very long to endure, and send you soon to us in help and prosperity; […]’
4.3 The Celtic hypothesis The issue of Celtic linguistic influence has attracted new attention in recent years, with claims that a number of Celtic structural features have survived in English (for a stateof-the-art report see Filppula et al. 2008). This has been supported by new historical and archaeological insights into the Anglo-Saxon conquest, and by preliminary studies of the genetic profile of Britain (cf. Filppula et al. 2008: 12–18). In this view, the invasion of Britain was a rather slow penetration of Germanic settlers into Celtic areas, with a continuity of Celtic settlement even in central parts of Britain. These historical facts support more intensive linguistic contacts, which is also compatible with the numerous surviving topographical names and with the proportion between Celts and Anglo-Saxons, which is estimated between 5:1 and as much as 50:1 (Filppula et al. 2008: 15). Traditionally, the imbalance between the large number of French loan words in English and the small number of medieval Celtic loans (see, however, Section 5.1.3) has been explained by the low status and small number of Celts in central England and the hostile relations between Celts and Anglo-Saxons. More recently, however, the scarcity of Celtic loan words has been linked to the relative status of the involved languages, with Celtic being the substrate language, and Old English the superstrate. Superstrate languages seem to mainly influence the lexicon of their substrate language(s) (cf. the large number of French loans after the Norman Conquest), while substrate languages affect their superstrate mainly on the grammatical level; accordingly, we should expect structural influence of the substrate Celtic languages on the superstrate Old English. This must have happened during the Celts’ shift to English within a few generations after the conquest, at least in the eastern and southern areas of Britain, a process involving imperfect learning. This scenario is compatible with the delayed emergence of most assumed Celtic features in ME texts, since they would first have entered a low variety of the spoken language and not the OE standard (cf. Section 5.1.2 for the similar delayed first attestation of Old Norse lexical loans).
32 Middle English: Language contact
5 Linguistic levels This section presents the main contact-induced changes with an emphasis on features which entered the standard language, though some purely dialectal features will also be looked at. Many of the changes resulting from contact with Old Norse are first recorded in Northern and Midland dialects, and slowly spread southward and partly into London as a result of dialect contact and demographic shifts. After a discussion of lexical borrowing, the most widespread and straightforward type of borrowing, the more controversial contact-induced changes on the structural levels will be dealt with. In many cases of assumed structural borrowing, language-internal, native factors have also been proposed as possible causes of change, especially by mainstream linguistics, while contact linguistics has, like earlier philological approaches, emphasized the importance of linguistic contact for linguistic change. Part of this ongoing controversy is linked to the question whether one looks for monocausal explanations of change or accepts that multiple causation, where foreign influence triggers or supports a native development, should also be considered as contact-induced change. The present survey will also list controversial cases of structural borrowing, since disregarding these would give a distorted picture of the dynamic research in contact linguistics. For reasons of space the often complex argumentation in these controversies cannot be addressed here.
5.1 Lexical borrowing The vocabulary of English mirrors the history of its linguistic contacts to a large extent. While the attested OE vocabulary amounts to about 25,000 to 30,000 lexemes, with only about 3% of mostly Latin borrowings, the Middle English Dictionary (Kurath et al. 1952– 2001) lists about 60,000 lexemes, 25% to 30% of which are loans. This dramatic increase is closely linked to the extensive borrowing from French (Fr.), Latin (Lt.) and Old Norse (ON). There are various reasons for these lexical borrowings, especially the high social and cultural prestige of French and Latin and the intimate contact between English and Old Norse as two closely related languages of similar status, but also lexical gaps, particularly in the case of cultural borrowing and the advance of learning. The extensive mixture of Germanic and Romance lexemes in Middle English affected the structure of semantic fields and led to numerous quasi-synonyms according to levels of formality, style or register, such as profound (Fr.) vs. deep or ill (ON) vs. sick. In many cases, one of the etymologically divergent quasi-synonyms eventually fell out of use or changed its meaning: thus ME em ‘uncle’ (< OE e¯am), led ‘person, nation’ (< OE le¯od(e)), etc. were finally replaced by the French loans uncle, people, while sky (ON ‘cloud’) took over part of the original meaning of native ME heven (OE ‘sky, heaven’); OE ce¯apman ‘merchant’ was in its general meaning replaced by Fr. merchant, but survived with a restricted meaning in chapman and as a personal name. Quite a number of ME loans survived only in dialects or fell out of use during the period. (For a detailed discussion of ME vocabulary, see Burnley 1992.) On the textual level, the distribution and frequency of loans in ME texts often differ according to text type and provenance: ON loans tend to occur more frequently in Northern and Midland texts, French loans in Southern texts; Latin or French loans are often more frequent in certain genres and in translations from these languages than in original English works.
511
512
IV Middle English
5.1.1 French and Latin loans A clear distinction between French and Latin loans is often problematic: Old French spelling is sometimes largely identical to that of the Latin source, while in other cases, the Latin spelling was changed to reflect integration into Middle English, such as allegory (ME allegorie < Lt. allegoria) or desk (ME deske < Medieval Lt. desca); finally, many apparent Latin or French words were actually coined in England on the basis of productive word-formation patterns using foreign lexical material (see Section 5.1.4). From a semantic point of view, Latin loans tend to be more learned than French ones, a reflection of the different status and functions of the two languages, though we also find double borrowing of cognates from French and Latin resulting in ME doublets, many of them surviving in modern English, such as (the first form is from French): loyal – legal; declension – declination; straight – strict (cf. Burnley 1992: 434). The huge number of French and Latin loans not only changed the structure of the English lexicon but also affected the phonological and morphological systems of English to some degree (see Section 5.2). With an estimated 10,000 items, French is the main source for ME lexical borrowing. However, borrowing started slowly, with fewer than 1,000 French loans attested before 1250, especially in “higher” fields like religion, the court, and literature (nativity, mercy; countess, chancellor, noble; story, rhyme). From about 1250 onwards, borrowing steadily increased, reaching its climax in the 14th century; i.e. most French loans entered English during the language shift from French to English. These later loans occur in many additional semantic areas where French was prestigious, such as administration, fashion, social life, food, medicine, and learning (see under [3]), but also in core areas of everyday life. Thus, we find French terms for tradesmen and artisans (butcher, barber, tailor), kinship terms (aunt, uncle, nephew, niece), including hybrids such as grand-father, granddaughter, and face as a body-part term. French loans are, however, rare in shipping and seafaring, as well as in farming (though farm itself is from French), which possibly reflects the lower prestige of these fields. (3)
Government, administration, law: government, state, parliament, treaty, tax, county; justice, court, crime, judge, complaint; to accuse, arrest, seize Military: army, navy, peace, enemy, battle, soldier Religion: religion, prayer, faith, temptation; divine, devout; to preach, repent Fashion: fashion, dress, coat, button, jewel, pearl; blue, brown Food: dinner, appetite, taste, bacon, venison, pork, sausage, salad; to roast, boil, fry Social life: music, conversation, chair, lamp, wardrobe
The above differentiation of two chronological strata of loans has traditionally been linked to different varieties of French, with the early group deriving from AngloNorman French, and the later as imported from the prestigious Central French. This seems supported by phonological correspondences in loans reflecting Anglo-Norman (AN) and Central French (CF) phonology, such as AN initial /k/ before /a/ against CF palatalization to /t∫/ as found in modern doublets like catch (< AN cachier) vs. chase (< CF chacer), cattle vs. chattel(s), but also in different spellings of the same word in Middle English, like canchelers (1066) vs. chanceleres (1300); calange (1225)
32 Middle English: Language contact vs. challenge (1300); equally, AN /w/ vs. CF /ɡ/ in originally Germanic loans, e.g. warden (1225) vs. guardian (1466); reward (1315) vs. regard (1430). This traditional view about two etymologically distinct chronological strata has, however, been criticized by Rothwell (1998); by providing text-based linguistic and historical arguments he shows that these phonological differences are not used consistently in ME texts and that many supposed continental loans are not only attested earlier in England than in France, but also differ semantically, facts he attributes to the ongoing lexical productivity of Anglo-French. Latin loans, predominantly taken from the written language, are relatively rare before the middle of the 14th century, but become frequent from then until the end of the period. They occur in all domains where Latin was widely used, such as religion, administration, medicine, and different branches of science, e.g., astronomy, botany, and chemistry. Some examples from these fields are aggravate, cause, client, contempt, conviction, desk, diaphragm, explicit, formal, legitimate, major, necessary, promote, psalm, substitute (see Burnley 1992: 432–433). Borrowing is also furthered by the great number of translations from Latin, where Latin words are frequently taken over, and by the increase in the production of English texts. Particularly important were the Wycliffian writings and Bible translations, which introduced more than 1,000 Latin words. In the late 14th and 15th centuries, Latinisms were increasingly used for stylistic reasons to make a writer’s style more elevated; this so-called “aureate diction” uses rare and unknown Latin words and is based on classical Latin rhetorics (Burnley 1992: 434). While traditionally borrowing from Latin has been seen as less extensive than from French, recent studies have found evidence for “a slight predominance of Latin” (see Vezzosi, Volume 2, Chapter 108).
5.1.2 Old Norse Most ON loans entered the spoken language of the Danelaw in the OE period, but first appeared only in ME texts. Estimates of their number greatly vary, but about 900 have survived in standard Present-day English (Kastovsky 2006: 223). Due to the sociohistorical context of the contact situation (see Section 4.1), ON loans occur in most areas of everyday life and most word classes: nouns (awe, birth, dirt, kettle, leg, sister), adjectives (awkward, flat, happy, weak), verbs (call, crawl, raise, want), but even function words such as pronouns (they, them, their, see Section 5.2.2; both, same), prepositions (till, fro) and conjunctions (though). About 600 more have survived in regional dialects, e.g. kirk ‘church’, benk ‘bench’, kist ‘chest’, trigg ‘true’. ON words either slowly replaced a native one, at least in the standard language, such as egg/ey†; sister/soster†, taken/ nimen†, or the ON and the native word became semantically differentiated resulting in doublets like skirt (< ON skyrta) vs. shirt (< OE scyrte), scatter vs. shatter; the latter two pairs also show the regular phonological correspondence between ON /sk/ and OE /ʃ/; further correspondences are ON /k/, /ɡ/ against OE /tʃ/, /j/, as in dike vs. ditch, kirk vs. church, give, get, whose voiced stop shows ON influence, replacing the OE forms with /j/ (cf. ME yeven/yiven ‘give’). The close relationship between Old English and Old Norse, however, makes some etymologies controversial. There are also some instances of semantic loans from Old Norse, such as earl, dream, whose meanings derive from ON jarl ‘chief’ and draumr ‘dream vision’, replacing the
513
514
IV Middle English original meanings ‘man, warrior’ and ‘joy’ of their OE cognates. (See further Dance, Volume 2, Chapter 110.)
5.1.3 Minor sources The number of Celtic loans first attested in Middle English is – with 83 items listed in the Middle English Dictionary (Kurath et al. 1952–2001) – larger than traditionally claimed (not to mention the numerous place-names), e.g. birling ‘a kind of boat’, commorth ‘a gathering’, laggen ‘make dirty’, mallok(e) ‘cursing’ (see Filppula et al. 2008: 128 and n. 58). Almost as many were borrowed from Low Dutch and Low German, such as trade, shore, mate, clock (den Otter 1990; cf. Hendricks, Volume 2, Chapter 105). Furthermore, there are some originally Greek, Arabic, and Persian words, mostly borrowed indirectly via other languages.
5.1.4 Word-formation The extensive borrowing of complex foreign lexemes from French and Latin (see Section 5.1.1) also affected ME word-formation and finally “led to a system with two derivational strata, a native and a foreign one” in late Middle or early Modern English, though it is still controversial when the foreign patterns became fully productive (Kastovsky 2006: 250–251; also Kastovsky, Chapter 9). Many OE prefixes died out or became unproductive in Middle English, while a large number of French and Latin prefixes were borrowed as part of individual complex lexemes, such as dis- (disobey, dispraise), in-/im- (ineffable, impossible), en- (engender), frequently together with their base forms obey, praise, etc. New formations first combined these foreign prefixes only with foreign bases, but in late Middle English we also find new hybrid formations with native bases, which seem to point to the beginning productivity of these patterns (e.g. enthrallen), see Burnley (1992: 447–449). The question of productivity is even more difficult with the numerous imported Romance suffixes: some of these never became fully productive, such as -ive, -ate, while others were used in new hybrid formations from the 14th century on, e.g. -age (barnage ‘infancy’, 1325), -ess (shepherdess, 14th century), -able (eatable, 1483). The development of phrasal verbs such as farenn forþ, commenn off, particularly from the 12th century on, is most likely due to the model of Old Norse, though French and native influence have also been proposed (cf. Fischer 1992: 386; Kastovsky 2006: 223). Phrasal verbs increasingly replaced the native prefixed verbs such as utgon, utfaran ‘go out’, often after a period of coexistence, also affecting English syntax.
5.2 Structural borrowing As stated above, foreign influence on structural changes is much more difficult to establish than lexical borrowing, and there has been a great deal of controversy in this area, especially with syntactic change. Discussions have been strongly influenced by underlying general assumptions, and mainstream historical linguistics has preferred languageinternal motivations of change over the working of language contact, while the latter has been favored by philologically or sociolinguistically oriented scholars (see Danchev and Kyto¨ 2001: 40). In recent years a slight change in attitude has been noticeable,
32 Middle English: Language contact compare, for example, the accounts of English historical syntax in Fischer (1992) and in Fischer and van der Wurff (2006), and multicausal explanations are increasingly accepted. For a comprehensive and balanced discussion of possible Celtic influence on ME grammar including areal linguistic considerations, see Filppula et al. (2008).
5.2.1 Phonology Contact-induced phonological change is clearly visible in the prosodic system of Middle English, where the Romance stress rule introduced through the numerous loans from French and Latin came to co-exist with the inherited Germanic stress rule with primary stress on the first syllable of the lexical root. While the latter assigns primary stress from the “left-hand word-edge”, the Romance rule “counts from the right-hand word-edge” (Lass 2006: 67–68; Ritt, Chapter 26). The coexistence of the two systems is frequently reflected in ME stress variation in Romance loans, such as dı´vers vs. dive´rse, but also in the stress assignment in derivationally related patterns such as hı´story – histo´rian/histo´ric (cf. Kastovsky 2006: 252). Foreign influence on ME segmental phonology is rather small. The ME phonemization of the voiced fricatives /v, z/, which were allophones of /f/ and /s/ in Old English and did not occur in initial position, is at least partly due to the new French/Latin loans with initial /v, z/, which led to the development of new minimal pairs like ferry vs. very (< Fr. verrai), seal vs. zeal (< Fr. ze`le); this development may have been facilitated by initial fricative voicing in native Southern ME dialects (still reflected in PDE fox vs. vixen) (cf. Lass 1992: 59, 2006: 62). In the vowel system, the borrowing of the two “distinctly non-Germanic” diphthongs /oi/ and /ui/ from French (joy, puison) is surprising, as they have no native parallels or sources (Lass 1992: 52).
5.2.2 Morphology One of the main structural differences between Old and Middle English is the attrition of inflectional morphology, which is linked to the typological change to a predominantly analytic language (for details see Thomason and Kaufman 1988: 277–282; Go¨rlach 1986). This development had already started in Old English as a result of Germanic stress, but its strong presence in early ME dialects of the Danelaw area points to the important role the Old Norse–English contact played in this change, though Celtic substratum influence via increased initial stress has been repeatedly proposed as well (Filppula et al. 2008: 71, 121–122). The extensive language-internal restructuring of ME syntax, such as word order changes and extension of periphrastic constructions has also to be seen against the background of this morphological attrition. A contact-induced change strongly affected the different forms of the personal pronoun of the third person plural, where the OE h-forms (hie, hira, him) were supplanted by the ON forms with þ/th- (cf. they, their, them). These forms slowly spread south to London, with Chaucer already having they, but still using native her(e), hem (Lass 1992: 120). The extension of one into a generic pronoun in the 15th century replacing ME man has been linked to the model of French on (Fischer and van der Wurff 2006: 116). Furthermore, French (and indirectly Latin) influence led to the pragmatically motivated extension of the native plural pronouns yee, you as polite singular address forms;
515
516
IV Middle English these were restricted to upper-class use around 1300 but became unmarked by the end of the period (Lass 2006: 96–97; Busse, Chapter 46). The widespread simplification and loss of verb inflection which started in the north and slowly spread south in Middle English was largely due to contact with Old Norse, though supported by native developments (Thomason and Kaufman 1988: 278–281). Direct ON influence is seen in a number of borrowed morphemes such as the Northern present participle -and (< ON -andi), the infinitive marker at, the present plural form are, and possibly the third person singular present -s (Thomason and Kaufman 1988: 293–295). A morphosyntactic feature found in ME Northern and Midland dialects is the “Northern Subject Rule”, i.e. the absence of present plural -s with verbs in contact position with subject pronouns, resulting in Children sings vs. We/you/they sing and plays. Both the rise of the construction in ME Northern dialects and a parallel construction in Celtic (Brythonic) languages point to a possible Celtic substratum origin (Filppula et al. 2008: 42–49).
5.2.3 Syntax Foreign syntactic influence may trigger or reinforce the development of a construction that already exists in embryo in the receiving language or for which a structural need arises, while constructions which are foreign to a language have been – though controversially – claimed to “have no possibility of gaining a footing in it” (Sørensen 1957: 133). Thus, contact-induced syntactic change is difficult to detect and even more difficult to prove, though certain criteria, such as type of contact, geographical distribution of features, and their occurrence in translated texts may help in the decision. Many of the proposed contact-induced syntactic changes are highly controversial, especially those of Old Norse and Celtic origin, while Latin and French influence have in general been more widely accepted. In most of the controversial cases discussed below, a polygenetic origin seems more likely than a monocausal explanation, both for the first appearance and the sometimes dramatic increase of a construction with a parallel in a contact language. Contact-induced syntactic changes seem to differ according to the status of the donor language, the type of contact, and language shift. The High languages Latin and French often seem to have served as direct or indirect models for English, while the – more controversial – Celtic substratum influence is more likely due to imperfect learning in the course of the Celts’ shift to English, and ON features result from the intimate contact between Old Norse and English in the Danelaw area (see Section 4.1 and Section 4.3). There are a few changes in the ME noun phrase where the rise and spread of a native construction seems to have been reinforced particularly by French influence. Among these is the postposition of adjectives after nouns, especially with learned French adjectives which often show French plural marking, as in other goodes temporels, though Latin influence may also have contributed in specific cases (Fischer 1992: 214; Sørensen 1957: 148). The replacement of the inflected genitive by the of-genitive phrase is undoubtedly of native origin, but its enormous increase in Middle English “may have been helped along by the parallel French construction with de” (Fischer 1992: 226; Fischer and van der Wurff 2006: 118).
32 Middle English: Language contact Even more foreign influence has been proposed for the verb phrase. A still unresolved controversy is the appearance of the historical present in later Middle English, being variously attributed to Latin or French influence or a popular native development (see Fischer 1992: 242–244; Sørensen 1957: 143). Even more hotly debated is the origin of the progressive form (see Denison 1993: 397–408), whose frequency in early Middle English is as low as in Old English, but strongly increases in late Middle English, particularly in northern texts. In the controversy about its origin and continuity from Old English, both formal and functional arguments have been brought forward. Fischer (1992: 253–255) acknowledges possible Latin influence on the OE form, but sees “far less evidence for the influence of French or – even less – Celtic on the development of the construction in Middle English” (253), while Fischer and van der Wurff (2006: 136) mention the possibility of a Celtic substratum. The most extensive argumentation in favor of a Celtic substratum of the progressive is found in Filppula et al. (2008: 59–72), who emphasize the close parallels between the English and Celtic (Brythonic) construction. (With this feature, a polygenetic origin also seems attractive, and at least the further functional extension of the progressive is a language-internal development [cf. Fischer and van der Wurff 2006: 136–137]). A similarly complex problem is the origin of “periphrastic do”, which is now predominantly seen as having originated in Middle English from the native full verb do, especially in its causative or anticipative function (Fischer and van der Wurff 2006: 154–155, Denison 1993: 256–264). Fischer (1992: 269) claims that possible Celtic or French influence is “no longer generally upheld” respectively “usually ruled out”, except for a possible indirect French influence on causative do (Fischer and van der Wurff 2006: 154; see also Denison 1993: 274–283). However, after extensive discussion of all available evidence, Filppula et al. (2008: 59) conclude that there is a strong case for Brythonic influence (see also Warner, Chapter 47). As for subordinate clauses, the ME rise of the “accusative-plus-infinitive” (as in We believe this to be wrong) from about 1400 is a widely accepted case of Latin syntactic influence, though its implementation was enabled by the establishment of VO order (see Fischer and van der Wurff 2006: 193–195; Sørensen 1957: 138-139). Finally, in the complex development of the different wh-relative pronouns, Latin and French seem to have strengthened a language-internal tendency (Fischer 1992: 299–301). In regard to word order, “verb-second” in declarative clauses generally decreases in Middle English, but is quite consistently used in 13th century northern texts from areas with heavy Scandinavian influence; however, the overall reduction of verb-second in Middle English has also been attributed directly or indirectly to French and Scandinavian influence, as has the change from OV to VO word order, though their further spread was certainly due to and linked with other native developments (Kroch and Taylor 1997; Fischer and van der Wurff 2006: 184–187, 193–194). The above discussion is in no way complete, and numerous other syntactic features have been claimed to be at least partly due to foreign influence, such as the early ME development of the for to + infinitive to express purpose (Scandinavian and French, cf. Danchev and Kyto¨ 2001; Fischer and van der Wurff 2006: 180), it-cleft sentences (Celtic substratum, cf. Filppula et al. 2008: 75–84), etc. The material presented here should, however, have shown the dynamic research carried out in this field in recent years.
517
518
IV Middle English
6 Conclusion The multiple language contacts in medieval England have left numerous traces on the different linguistic levels of English and are at least partly responsible for the dramatic changes in the vocabulary and grammar of Middle English. Most of the linguistic results of the early contacts with speakers of Celtic and Old Norse only surface in texts from the ME period, while those due to French and Latin are more directly observable. The dramatic effects of borrowing on the lexicon of Middle English are rather uncontroversial, whereas contact-induced changes on the structural levels are more controversial, since a foreign feature may only have reinforced a linguistic tendency already present in English. In any case, contact-induced change has to be approached in the sociohistorical context of the various languages in contact, and both extralinguistic and linguistic factors must be taken into account in order to arrive at a balanced view of the role which extended language contact has played in the extensive restructuring of Middle English.
7 References Bailey, Charles-James N. and Karl Maroldt. 1977. The French lineage of English. In: Ju¨rgen M. Meisel (ed.), Langues en Contact – Pidgins – Creoles – Languages in Contact, 21–53. Tu¨bingen: Narr. Blake, Norman (ed.). 1992. The Cambridge History of the English Language. Vol. II. 1066–1476. Cambridge: Cambridge University Press. Burnley, David. 1992. Lexis and semantics. In: Blake (ed.), 409–499. Danchev, Andrei and Merja Kyto¨. 2001. The Middle English “for to + infinitive” construction: A twofold contact phenomenon? In: Dieter Kastovsky and Arthur Mettinger (eds.), Language Contact in the History of English, 35–55. Frankfurt: Lang. Dawson, Hope C. 2003. Defining the outcome of language contact: Old English and Old Norse. Ohio State University Working Papers in Linguistics 57: 40–57. den Otter, A. G. 1990. Lekker scrabbling: Discovery and exploration of once-Dutch words in the online Oxford English Dictionary. English Studies 71: 261–271. Denison, David. 1993. English Historical Syntax: Verbal Constructions. London: Longman. Filppula, Markku, Juhani Klemola, and Heli Paulasto. 2008. English and Celtic in Contact. New York: Routledge. Fischer, Olga. 1992. Syntax. In: Blake (ed.), 207–408. Fischer, Olga and Wim van der Wurff. 2006. Syntax. In: Hogg and Denison (eds.), 109–198. Go¨rlach, Manfred. 1986. Middle English – a creole? In: Dieter Kastovsky and A. Szwedek (eds.), Linguistics across Historical and Geographical Boundaries. In Honour of Jacek Fisiak, 329–344. Berlin/New York: Mouton de Gruyter. Halmari, Helen and Timothy Regetz. 2011. Syntactic aspects of code-switching in Oxford, MS Bodley 649. In: Schendl and Wright (eds.), 115–153. Hickey, Raymond (ed.). 2010. The Handbook of Language Contact. Malden, MA/Oxford: WileyBlackwell. Hogg, Richard and David Denison (eds.). 2006. A History of the English Language. Cambridge: Cambridge University Press. Kastovsky, Dieter. 2006. Vocabulary. In: Hogg and Denison (eds.), 199–270. Kibbee, Douglas A. 1991. For to Speke Frenche Trewely. The French Language in England, 1000– 1600: Its Status, Description and Instruction. Amsterdam/Philadelphia: John Benjamins. Kroch, Anthony and Ann Taylor. 1997. Verb movement in Old and Middle English: Dialect variation and language contact. In: Ans van Kemenade and Nigel Vincent (eds.), Parameters of Morphosyntactic Change, 297–325. Cambridge: Cambridge University Press. Kurath, Hans, Sherman M. Kuhn, John Reidy, and Robert E. Lewis. 1952–2001. Middle English Dictionary. Ann Arbor: University of Michigan Press. http://quod.lib.umich.edu/m/med/
33 Middle English: Standardization
519
Lass, Roger. 1992. Phonology and morphology. In: Blake (ed.), 23–155. Lass, Roger. 2006. Phonology and morphology. In: Hogg and Denison (eds.), 43–108. Rothwell, William. 1998. Arrivals and departures: The adoption of French terminology into Middle English. English Studies 79: 144–165. Schendl, Herbert. 2002a. Mixed-language texts as data and evidence in English historical linguistics. In: Donka Minkova and Robert Stockwell (eds.), Studies in the History of the English Language. A Millennial Perspective, 51–78. Berlin/New York: Mouton de Gruyter. Schendl, Herbert. 2002b. Code-choice and code-switching in some early fifteenth-century letters. In: Peter J. Lucas and Angela M. Lucas (eds.), Middle English from Tongue to Text, 247–262. Frankfurt: Lang. Schendl, Herbert and Laura Wright (eds.). 2011. Code-Switching in Early English. Berlin/Boston: De Gruyter Mouton. Sørensen, Knud. 1957. Latin influence on English syntax. Travaux du cercle linguistique de Copenhague 11: 131–155. Thomason, Sarah G. 2001. Language Contact. An Introduction. Edinburgh: Edinburgh University Press. Thomason, Sarah G. and Terrence Kaufman. 1988. Language Contact, Creolization, and Genetic Linguistics. Berkeley: University of California Press. Townend, Matthew. 2006. Contacts and conflicts: Latin, Norse, and French. In: Lynda Mugglestone (ed.), The Oxford History of English, 61–85. Oxford: Oxford University Press. Wenzel, Siegfried. 1994. Macaronic Sermons. Bilingualism and Preaching in Late-medieval England. Ann Arbor: University of Michigan Press. Wright, Laura. 2011. On variation in medieval mixed-language business writing. In: Schendl and Wright (eds.), 191–218.
Herbert Schendl, Vienna (Austria)
33 Middle English: Standardization 1 2 3 4 5 6
Introduction Standardization(s) and standards Pragmatically focused Middle English varieties From external to internal standards Summary References
Abstract Middle English is usually seen as the period of English which does not have a supraregional standard because French and Latin fulfil this function. However, at least for the 14th and 15th centuries fairly standardized varieties with dialectally bleached forms may be identified. Yet such varieties may only be regarded as “focused” because codification as the sufficient condition for a standard is still absent. By the 15th century, when English is more extensively used in writing, the norm was mainly that of discourse traditions which were taken over from the French and Latin models as external standards. These provided English with means not only to communicate over a geographically wider area, but also to produce texts in an increasing number of registers as these external standards become internalized in English. Middle English standardizations are hence Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 519–533
520
IV Middle English discussed here as a minimization of variation for supra-local use and as maximation of variability to widen the stylistic functions of the vernacular.
1 Introduction Middle English is traditionally considered to be the standardless period of English. The absence of a standard is all the more a defining quality of Middle English because what we regard to be the “beginning of Middle English” coalesces with the disappearance of the so-called “West-Saxon Literary Standard” (cf. Kornexl, Chapter 24), so that Roy Liuzza (2000: 145) could claim that “the end of […] scribal training [in the West-Saxon standard] was the beginning of ‘Middle English’ ”. Moreover, in the last century of the Middle English period the vernacular is generally supposed to be under way towards Modern Standard English (cf. Moessner, Chapter 44). Situated between the disappearing of one written standard and the slow emerging of a new standard, Middle English therefore “exhibits by far the greatest diversity in written language of any period before or since” (Milroy 1992: 156). The reasons for this linguistic scenario lie in the specific language situation resulting from the cultural impact of the Norman Conquest on the one end and the rapidly increasing use of English in writing for various purposes on the other. This article is structured in such a way that Section 2 discusses different concepts of the notions “standardization” and “standard” as they are applicable to Middle English. Section 3 subsequently discusses “standardized varieties” of English which have been identified mainly on the ground of unified spellings, and, in the 14th and 15th centuries, also in terms of the spread of dialectally bleached forms which could function supralocally. Section 4 then concentrates on the extensive and intensive elaboration of written English in the 14th and 15th centuries. Section 5 gives a brief summary of the main aspects addressed in the article.
2 Standardization(s) and standards In general the idea of standardization is based on a concept of fixing and fixity of what would otherwise vary more or less freely. As the prime – and for a long time sole – means of fixing (a) language, writing potentially produces such models that minimize variation (cf. Percy, Chapter 63). However, writing is a necessary, but not a sufficient, condition for such a specific variety to emerge because it seems to take further sociopragmatic motivations and textualization efforts for a standard variety to become established. Under “socio-pragmatic” shall be subsumed such standardizing moves that modify written language to serve a wider geographical range of communication. By “textualization efforts” I refer to orientation towards standards already established by existing literate models that will be called “discourse traditions”, a notion specifically apt to account for the unique literary communicative space of the period under consideration. As shall be seen, the former moves reduce variation while the latter efforts enhance variability.
2.1 Standardization for de-localized communication Sandved (1981: 31) programmatically states that “the term ‘standard language’ is essentially a socio-linguistic term” applying to “any form of written English […]
33 Middle English: Standardization regarded as a model worthy of imitation by people belonging outside the geographical area or social group within which this variety arose” (italics in the original). Inferring a social gain – such as “prestige” – for people “outside […] a social group” who imitate the variety in question is as simple as it is hard to substantiate. In contrast, one may assume, it is much easier to identify the gain of imitating a variety used in a geographically different area. Although Sandved does not explicitly say so, he seems generally to have spelling in mind, and in particular that spelling which was seemingly developed by “the Chancery” in the 15th century and copied in the province. The readiest assumption is that 15th-century “imitators from outside” had realized the communicative profit of using a spelling norm which evens out dialectal differences. This idea is further elaborated by Jeremy Smith (1992, 1996, 2000), who has also heuristically specified the notion of a “standard variety” – with standard in adjectival use – in that he distinguishes “standardized” from “standard” phenomena. He suggests the notions “standard” (ADJ) or “fixed” to relate to a prescriptive, codified variety, and “standardized” or “focussed” to relate to a norm at which speakers or writers (may) aim (Smith 1996: 65–66). According to Smith (2000: 136), in the Late Middle English period “standardization” brings about the “evolution of […] a colourless regional usage”, which, in his terminology, functions as a “focus” rather than a “standard”. With specific regard to the 15th-century adoption of “Chancery English” spelling, he states that this catered to the “demand for more broadly diatopic communicative efficiency” (2000: 136) and thus is pragmatic in nature. All in all, much scholarly effort has been spent on identifying some contemporary organizing rationale in writing the vernacular. Spelling looms so large in these efforts that the establishment of other non-regional norms, such as evened-out dialectal differences on the level of grammar, are usually fused with spelling norms proper, as is the case in Samuels’s “Types I–IV” (cf. Samuels 1989 [1963]; Smith 1996: 66–73; cf. also Section 3.2 and Section 3.3 below). To these belong the spread of the present participle ending -ing and the th-forms of the third-person plural pronouns originally borrowed from Old Norse. However, in the last decade or so the research on further Middle English grammatical developments that inform Modern English and subsequently Modern Standard English has considerably increased, as will be seen in Section 3.4. Moreover, the heuristic scope has been widened by taking into acccount the discursive environment in which historical linguistic developments in general and those towards Modern Standard English in particular are couched (see the contributions in Diller and Go¨rlach 2001).
2.2 Standardization through elaboration In a publication on the development of the scientific register in late Middle and early Modern English, Irma Taavitsainen (2001: 185) states that for scientific writing in English the “foreign model of Graeco-Roman writing provided a standard, which can be seen as the norm towards which vernacular writing tended”, and that this norm influenced “both the macrostructures of discourse and the microlevel of linguistic features”. These observations are true not only for English scientific writing, but may be extended to other kinds of discourse traditions. The notion “discourse tradition” as it has been suggested by Peter Koch and Wulf Oesterreicher (1994: 589) loosely covers “Textsorten, Gattungen, Stilrichtungen, Gespra¨chsformen”(‘text types, genres, styles, conversational forms’). Thus conceived, the notion is heuristically profitable because
521
522
IV Middle English it helps to conjoin two important aspects which have to be taken into consideration when studying the reestablishment of English in writing: (1) the fact that “textualization” (Verschriftlichung) of a language never happens in terms of the “entire” language, but proceeds along the lines of “discourse traditions of distance” (Koch and Oesterreicher 1994: 594; for the notion “(language of) distance”, see Koch and Oesterreicher 1985); (2) that the massive “(re-)textualization” of English unfolded in a communicative space in which the literary languages French and Latin had their established functions and manifested stable “discourse traditions of distance”. As discourse traditions are, in principle, independent of an individual language (Koch and Oesterreicher 1994: 589), they may “travel”, as it were, between languages and thus, in Taavitsainen’s words, act as models or norms for the “macrostructures of discourse and the microlevel of linguistic features” (Taavitsainen 2001: 185) in different languages. A prime example for such a discourse tradition would be the so-called “curial style” which has been identified by Burnley in insular Latin, French and English writings (cf. Burnley 1986, 1989, 2001). The later Middle English striving to match the norms already in existence in the other literate languages may, in Einar Haugen’s (1966: 933) terms, be identified as “elaboration” (Ausbau). “Elaboration” figures as one of the four processes consitutive for a standard language as in Table 33.1: Table 33.1: Haugen’s (1966: 933) standardization parameters Society Language
Form Selection Codification
Function Acceptance Elaboration
“Elaboration” enables a language to serve a “maximal variation in function”, while “codification” aims at “minimal variation in form” (Haugen 1966: 931; italics in the original). If we disregard the prescriptive artes dictaminis in Latin and French, which also helped, for instance, the Pastons to compose their “private” letters in the 15th century (Nevalainen 2004; Schaefer 1996), codified prescription as we know it from the 18th century onwards (see Crowley, Chapter 61), can, however, not be found in the pre-modern eras of English. As to “selection of some kind of a model from which the norm may be derived” (Haugen 1966: 932; italics in the original), both the Middle English following of norms that serve the de-localization of English, such as spelling and certain grammatical forms, and those that already existed in terms of discourse traditions have to be taken into account. Both are, indeed, “socially” relevant, yet the first compliance with norms is largely pragmatic in nature, and indeed aims at a formal reduction of form. The second, in turn, aims at preserving the function of the already accepted norms in the other literary languages in order to elaborate English through imitating the French (and Latin) models. In this latter kind of standardization, the range of the native linguistic inventory is increased by new forms adding further “variability and expressiveness” (Rissanen 2006: 136). To grasp this increase of forms and means, Koch and Oesterreicher (1994: 589) prefer to speak of “intensive elaboration”, and to call the spread over a larger range of communicative functions “extensive elaboration” (italics added).
33 Middle English: Standardization
3 Pragmatically focused Middle English varieties In terms of specific varieties, various Middle English “standards” have been discussed in the literature. As has been indicated before, these “standards” are largely restricted to (otherwise not codified) spelling norms. Although the first of these varieties to be considered in the present section is relatively early and locally quite restricted, far-reaching claims have been deduced from its existence. The next group of varieties dates into the 14th century and is constituted by scribal “standards” that also comprise the use of specific grammatical forms. The final variety to be treated in this section is the 15th-century “Chancery English” or “Chancery Standard”.
3.1 The “AB Language” From the first quarter of the 13th century, two Middle English manuscripts have come down to us which contain texts with a “virtually identical language of unusual internal consistency” (Black 1999: 157). One of these texts is a version of the Ancrene Wisse (“Rule for Anchoresses”) preserved in MS Cambridge, CCC, 402, the other is a group of texts preserved in MS Oxford, Bodley 34, usually refered to in the literature as the Katherine Group. Tolkien has dubbed the former text “A” and the latter group of texts “B” (Black 1999: 157). In his edition of the Life of Juliana from MS Oxford (Bodley 34) d’Ardenne (1961 [1936]: 178) generally characterizes the “AB language” as containing “much that was ancient […], and indeed, it may be suspected, purely literary and traditional”. As a matter of fact, when reading these texts it is much more helpful to have a solid knowledge of Old than of Middle English. Yet d’Ardenne mainly sees “close relations […] between the orthography [of “AB”] and that of Old English” which ultimately leads him to the conclusion that “AB” must have directly descended from (West Saxon) Old English (d’Ardenne 1961 [1936]: 178). As Black (1999) has illustrated, such a direct descent is hard to substantiate. Thus, Jeremy Smith (2000) attributed to “AB” at best the quality of a focused variety, and Black (1999: 166) is even more uncompromising by suggesting that “A” and “B” may be the product of one single scribe with an extraordinarily “systematic mind”. The research on “AB” is illustrative of some general features that inform the research on standard English and English standardization on a larger scale. In his attempt to account for the relatively uniform spelling of “AB”, Tolkien had already surmised that the “consistency and individuality of the spelling […] suggests obedience to some school or authority” (Tolkien 1929; quoted in Black 1999: 157). Black (1999: 156) criticizes such “conjectures” (proliferated in textbooks) which work “on the assumption that the [“AB”] language, if it was a standard, must have been developed at a sizeable centre”. The acquisition of an orthography, we tend to infer, is the consequence of formal schooling. Thus, wherever there is such a norm detectable, there must have been a concrete institution that has formally passed this norm on. The most prominent example is that of the so-called “Chancery Standard”, as will be seen in Section 3.3. However, if Black’s hypothesis of a single “AB” scribe is tenable, that of a “school” in which that uniform spelling was trained is void. Last but not least, there is d’Ardennes’s (1961 [1936]: 178) “descent hypothesis”: that is, the claim that there is an unbroken “Anglo-Saxon lineage” of “AB” as “literary idiom” which has its ancestry in the 9th century Old English glosses of the Vespasian
523
524
IV Middle English Psalter. D’Ardenne thus positions “AB” in a deep-down, unbroken, teleological history, which, by implication, should hold for the entire history of the English language (and literature). Although James Milroy (1996: 180) does not directly refer to d’Ardenne (but to Ancrene Wisse), he criticizes such narratives of continuity as helping “to establish a long history for the language, which can be seen retrospectively as part of the process of legitimization of modern standard English as a powerful symbol of English nationality”. In order to substantiate that English prose had never ceased, Chambers (1932) in his famous piece On the Continuity of English Prose from Alfred to More and his School also claims Ancrene Wisse to be a link between Old and Early Modern English: “[Ancrene Wisse], even if we date it c. 1200, takes us back to a period when the tradition of pre-Conquest prose was still alive – and its forward links are even more important than its links with the past” (Chambers 1932: xcviii). In the terminology introduced earlier in the present article, we could translate Chambers’s postulate into saying that “AB” continues and preserves an Old English discourse tradition and passes it on up to the early Modern period. Yet this would be stretching the notion of “discourse tradition” too far and over too long a period in which we would have to surmise the respective tradition to have invisibly hibernated. In sum, “AB” should be regarded as “one of the local [South-west Midland] attempts to reorganize the traditional spelling of the area” with traces of the persistence of “Old English traditions of text-production” (Smith 2000: 130). Any further claims, such as the idea that “AB” is evidence of an institutional effort towards standardization or a link connecting Old and Modern English prose traditions, feed narratives of unbroken continuity. “AB” remains a solitaire in the earlier period of Middle English, when text production in the indigenous vernacular was quantitatively the absolute exception. Yet this situation clearly changed in the 14th and 15th centuries, as will be seen in the following section.
3.2 Samuels’s “Types I–III” Within the research project on Middle English dialects that has finally led to the Linguistic Atlas of Late Middle English (LALME) and Linguistic Atlas of Early Middle English (LAEME), M. L. Samuels in 1963 opened the following prospect of “provid[ing] us with a frame of reference for isolating and classifying those types of language that are less obviously dialectal, and can thus cast light on the probable sources of the written standard English that appears in the fifteenth century” (Samuels 1989 [1963]: 66). In his article Samuels identifies four such types. At present I will refer only to “Types I-III” because to “Type IV”, known as “Chancery English”, some scholars have attributed so much importance that it is worthwhile treating it separately. Samuels (1989 [1963]: 67) characterizes “Type I” as “difficult to localise”, but finally suggests that it is “a standard literary language based on the dialects of the Central Midland counties”. It is found in Wycliffite texts, which were produced in large quantities until 1430. “Type II”, which chronologically precedes Type I, is located by Samuels (1989 [1963]: 70) in the “greater London area”. Eight 14th century manuscripts – among them (part of) the Auchinleck Manuscript (c.1330) – show features of this type. “Type III” is considered by Samuels “as representative of London English of 1400” (1989 [1963]: 71). Among others it is the language of Chaucer, of one of the Piers Plowman manuscripts, and of Hoccleve (Samuels 1989 [1963]: 70). As Type III
33 Middle English: Standardization is considerably different from the earlier London Type II, Samuels (1989 [1963]: 70) concludes “that the London dialect changed suddenly and radically in the fourteenth century”. This change is usually attributed to increasing migration from the Central Midlands and East Anglia to London (Benskin 1991: 77). Benskin characterizes the linguistic result of this continuing migration as a “largely colonial dialect” developing “from the common core variants of diverse immigrant speech” (Benskin 1991: 78). This de-regionalized variety is thus not brought about by the scribal evening-out of dialectal variation in writing, but rather reflects in writing a kind of koine¨ization that has evolved in the spoken language of the capital. Both the last quotation from Samuels and Benskin’s characterization of Type III make it clear that these “Types” have been identified not only in terms of relatively unified sets of spelling. Thus, the London Type II has the “traditional” present participle endings -ande/ -ende/ -inde in the earlier 14th century, while the Chaucer manuscripts in Type III have -yng (Samuels 1989 [1963]: 70), which testifies to a further supra-localization of the -ing variant. Similarly Chaucerian they vs. Type II þai/hij testifies to the ultimate exclusion of older hij (< OE hi) from the norm in favor of the orginally Old Norse form. These remarks are in order because in the literature the “Types” suggested by Samuels are mostly treated as “mere” spelling foci. Thus, Smith somewhat confusingly treats the “Types” in one publication in a chapter on “Orthography” (Smith 1992: 55–57), in another publication in a chapter on “Change in writing system” (Smith 1996: 68–71). Nevertheless, Smith also sees in these focused varieties the interplay of the growing “literacy in the vernacular” and the attempt to overcome the “communicative dysfunctionality of massive diatopic variation” (Smith 2000: 135), be it graphic or morphological in nature. The peak of this supra-regional “bleaching” was allegedly reached in the 15th century with the so-called “Chancery Standard”, to which we turn next.
3.3 “Chancery English” For various reasons Samuels’s (1989 [1963]: 71) “Type IV” which he has called “Chancery Standard” deserves a subsection of its own. This type “consists of that flood of government documents that starts in the years following 1430 […], and it is this type, not its predecessors in London English, that is the basis for modern written English” (Samuels 1989 [1963]: 71). Again the criteria for identifying this “standard” are spelling styles and the use of specific (grammatical) forms. Regarding “Chancery English” as the direct ancestor of “modern written English”, if not “Modern Standard English”, has very much become the received wisdom, and one may surmise that Type IV exercises such an attraction for scholars studying the beginnings of Modern Standard English because it can be connected directly to a powerful institution – and ultimately to the king. Apart from Samuels who first proposed the name, various publications by John H. Fisher (among them those gathered in Fisher 1996) have contributed to assigning “Chancery English” this eminent place. However, this claim has meanwhile come under critical scrutiny. In his 1977 article on “Chancery English and the Emergence of Standard Written English” Fisher relativizes the link between “Chancery English” and Modern Standard English by conceding that “Modern English is not Chancery English”; yet he postulates that “Chancery English of the early fifteenth century is the starting point for this evolution [towards Modern English], and it has left an indelible impression upon the
525
526
IV Middle English grammar, spelling, and idiom of Modern English” (Fisher 1996 [1977]: 64). Although administrative written English of the 15th century may indeed have been an important path in the geographical spread of linguistic phenomena (cf. Section 4.2 below), scholars have spotted weak points in Fisher’s reasoning which very much downsize the all-encompassing claim for “Chancery Standard”. For one thing, the very notion “Chancery Standard” is regarded by Benskin (1991: 79) as “a considerable misnomer” possibly induced by the German Meißener Kanzleisprache, which was Martin Luther’s model variety. As Benskin (1991: 79) and Ormrod (2003: 785) have made clear, as of 1417 the English documents under consideration were issued by the Signet, that is, the King’s Secretary’s Office and later by the Privy Seal Office. These are, however, not to be identified with “the Chancery”. Moreover, Benskin states that the relatively small number of clerks working in the king’s offices were not trained “in the households of Chancery” (Benskin 2004: 35, 1991: 81). And yet another misunderstanding which Benskin uncovers is that of claiming the scribes to have been “subject to deliberate control” (Benskin 2004: 33). Last but not least, Ormrod rejects that Henry V might have had any conscious language policy which could have contributed to the establishing of English in general and “Chancery English” in particular as “the language of the administration” (e.g. Fisher 1996 [1992] or Richardson 1980; cf. my criticism in Schaefer 2008). Thus, both philological and historical research invalidate the claim that Henry V ever made “a dynamic and decisive contribution to the wider development of standard English during the fifteenth century” (Ormrod 2003: 785). Moreover, this “language of the administration” lacks Haugen’s main feature of a standard, namely that of “maximal variation in function” (Haugen 1966: 931), a quality which Modern Standard English, of course, has obtained. Therefore, we must agree with Laura Wright’s statement that it is “[…] not adequate to suggest that this diversity of function [of Modern Standard English] could have arisen solely from Chancery documents” (Wright 1994: 113). Apart from the functional limitations of “administrative English” in the 15th century, another important aspect about “the language of the administration” has to be kept in mind. English was not the language of the administration, but rather the minor partner of French and Latin. As Benskin (2004: 38) unmistakably states: “Chancery Standard was Latin, and save for nine years during the Commonwealth, it remained so until 1731, when for official purposes it was abolished altogether by Act of Parliament”. In terms of the factor “discourse tradition”, this emphasizes that putting “Chancery English” into place means situating it in the multilingual discourse community concerned with the production of adminstrative texts. Further detailed research into that “flood of [English] government documents that starts in the years following 1430” (Samuels 1989 [1963]: 71) may, in other words, be really profitable if we regard “Chancery English” as evidence for the attempts to provide the English language with new means of expression – albeit within the obvious functional limits of that variety. Studies such as Rissanen’s (e.g. 2000, 2002), give interesting evidence of English as taking shape in a trilingual workshop. This idea shall be further pursued in Section 4 below. At present we need to return once more to Haugen’s process of “acceptance” as an integral part of standardization, as we look into supra-localized phenomena which are not covered by Samuels’s four types.
33 Middle English: Standardization
3.4 Supra-localizated linguistic phenomena “from below” Section 3.2 mentioned that Samuels’s “Types I–IV” also include grammatical phenomena such as the southward spread of Old Norse th- forms at the expense of the indigenous h- forms for the personal and the possessive pronoun of the third person plural. With the knowledge as to what ultimately made it into Modern English, such forms can be considered as a development from dialectally restricted to “standardized” forms in terms of options for the process of “selection” and subsequent “acceptance”. In contrast to phenomena whose status changed from “dialectally restricted” to “supralocal”, other grammatical phenomena that must have contributed to de-localized communication deserve our attention. Looking into such phenomena gears the perspective away from “standard” or “standardized” varieties as a product and concentrates on linguistic developments that ultimately also inform the Modern standard. To identify steps in this process, Nevalainen has suggested a very useful heuristic concept that helps to translate Haugen’s parameter of “acceptance” into a gradual process of changing communicative practice (cf. Nevalainen 2000: 333) which results in supra-localized use. I shall name only a small selection of such grammatical phenomena. The first is -ly as the de-adjectival adverbial marker going back to adverbially used Old English adjectives with -lic. According to Nevalainen’s (2006: 124) corpus data, the pattern was already established in the period between 1350 and 1420, and in Modern English this has become the systematic norm. The second is the distribution of “Genitive vs. the of-construction” (Nevalainen 2006: 119–20). This variation is, of course, the result of the typological change of English from (prevalently) synthetic to (prevalently) analytic triggered in the transition from Old to Middle English (cf. Wełna, Chapter 27; Smith, Chapter 28; Trotter, Volume 2, Chapter 114). Yet while all other case markers or remnants of them quickly vanished in Middle English, genitive -s has been preserved to date. According to the results discussed by Nevalainen (2006: 119–20), the use of the prepositional phrase rapidly outstrips that of the inflected form in the later 14th century. Another field for which we must infer both change “from below” and supralocalizaton is that of the Scandinavian lexical loans (cf. Dance, Volume 2, Chapter 110; Schendl, Chapter 32). While the late medieval lexical influx from French and Latin is repeatedly discussed in the literature, the Scandinavian lexical contributions have received much less attention when discussing the Late Middle English input to Modern (Standard) English. Besides the obvious lexical contributions from Old Norse like sky or skirt, Old Norse-derived Modern English nouns such as fellow or law and verbs such as call and dwell are today register-neutral, and probably already were so in Chaucer’s time (Smith 1992: 64). The fact that such “everyday” words have been borrowed at all is usually attributed to the adstrate contact situation in the Danelaw in the later 9th and during the 10th century. Moreover, the fact that these words spread to the South is attributed to the general Anglian influence on London English (cf. Thomason and Kaufman 1988: 302–304), which has already been considered for Samuels’s Type III (see above, Section 3.2). As the etymological project of the LAEME is not yet completed (in 2010), we can only assume how the reduction of variation beween Scandinavian loans and their English semantic equivalents proceded. Here things may turn out to be much more complicated than we would want to have them. Thus, a brief check of the entries for the verb call in the Middle English Dictionary (MED) (Kurath et al.
527
528
IV Middle English 1952–2001) gives a quotation from St. Margarete, one of the early 13th century texts from the Western Midlands in the “AB Language” (see Section 3.1 above), as the oldest entry in the synonymous twin-collocation clepien & callen (MED, s.v. callen, def. 1c). About 180 years later, Chaucer still uses clepe alongside calle. In a later 13th century version of the St. Margaret legend also from the Western Midlands, felaw (from ON felagi; MED, s.v. felau(e), def. 1a) seems to have the same semantic range as fere (from OE gefera; MED, s.v. fere, def. 1a), and so do both forms a hundred years later, for instance, in Gower’s Confessio Amantis (MED, s.v. felau(e), def. 1a and fere, def. 1a). These few samples, which are anything but representative, already show, for one thing, that by the end of the 14th century the competition between Old Norse and indigenous English forms is not yet decided in London usage. Moreover, where both are used, they seem to be on an equal footing registerwise. The contact situation between English and French (cf. Skaffari, Volume 2, Chapter 106) but also with Latin (cf. Vezzosi, Volume 2, Chapter 108), was, in turn, very different, and so was, grosso modo, the turnout of loans from French (and Latin) with regard to lexical variation in terms of nouns, verbs and adjectives. Moreover, by way of borrowing and calquing, new prepositions, adverbs and conjunctions were established in English, considerably increasing its syntactic versatility, as will be illustrated in the following section.
4 From external to internal standards With the exception of “AB” as an extremely localized “standard”, the standardizations discussed in Section 3 concentrated on the striving for supra-localizations in spelling and specific grammatical forms that seem to have mainly served pragmatic purposes in that they made transdialectal communication in writing easier. Nevertheless, “prestige” as the motive for following a specific model has been adduced more or less prominently to “explain” the respective “standardization”. “Prestige” is, in turn, also often put forward when it comes to assessing the relation between Middle English on the one side and Latin and French on the other. Latin and French have been identified as fulfilling “some of the functions of a standard language in Britain” (Go¨rlach 1990a [1988]: 17), while Middle English is considered “a typical case of a language of low prestige” (Go¨rlach 1990b [1986]: 74) at the time. According to Fisher (1996 [1992]: 22), matters dramatically changed after Henry V had written his first missive in English in August of 1417, but this certainly draws too simple a picture of how and why English in the 15th century started to develop into a multi-purpose language in writing. The so-called “Statute of Pleading” of 1362 – which is so often judged as having considerably raised the status of English – indeed mandates that English be the (spoken) language of pleading at court. Yet it does so in French – and in the complicated left-branching syntax of “courtly style” – because, at the time, French was the language in which royal statutes where published; and it also regulates that the pleas be “entered and enrolled” in Latin, because this was the language of permanent record. This throws light on a functional distribution of the available languages: if we want to interpret this distribution in terms of “prestige”, it may be wiser to deduce “prestige” from the functions of the discourse traditions rather than from the individual languages as such. The functional distribution, in turn, can be attributed to the way in which Latin and French were already “elaborated” to serve various functions in
33 Middle English: Standardization writing, and simultaneously to the fact that this was not yet the case for English. In short, it could be reasoned that the “restricted use” of Middle English indicates “that means of expressing complicated syntactic patterns were not sufficiently well developed” (Go¨rlach 1990a [1988]: 21). This very general assessment should, however, be specified in order to be able to retrace which steps improved the syntactic capabilities of English. As has been stated before, languages, for one thing, enter the literary realm along the paths of discourse traditions (Koch and Oesterreicher 1994: 594; see Section 2.2 above). Moreover, we need to take into account that written discourse traditions were obviously firmly established in England in both Latin and French. In view of these observations we may infer that those discourse traditions realized in Latin and French constituted norms to be aimed at in English as well. This is conceptualized here as the adoption of “external standards” in English which, when they have finally left their orginal discourse-traditional habitat, become “internal standards”. As substantial research within this heuristic framework is only slowly beginning, the following selective remarks have to suffice in order to illustrate which results this approach to standardization may still provide in the future.
4.1 Elaboration of open-class lexis As has been indicated, on a large scale French may be considered the supra-local standard in the 13th and 14th centuries (Burnley 1989: 35), because it could serve a wide range of functions. In that sense Burnley (1989: 36) has stated that French is “a source of supra-dialectal lexis” in English. Instead of generally associating French in England with “prestige” and identifying this as the reason for lexical borrowings into English, we should rather suppose that the taking-over of lexical material from French (and Latin) to English was, once again, a matter of “specific-purpose” (con-)texts, and thus of discourse-traditional norm compliance. Subsequently such borrowings must have lost their “technical” marking and could simply serve supra-local communication. By the time of Chaucer, early borrowings – such as the often-quoted prison – which mainly came from the field of administration, have obviously become integral unmarked elements of the “systematic norm” (cf., e.g., Smith 1992: 61–62). Borrowings of this kind need, however, to be distinguished from that large number of loans which later inform the lexicon of Modern English and which obviously entered the linguistic system from written French or Latin (cf. Go¨rlach 1990b [1986]: 74) in the late Middle English period. Again it seems a moot question to ask how “prestigious” these newly arriving loans actually were. In addition, one must be skeptical about claims that authors like Chaucer intentionally borrowed a large number of words from French and Latin to give English “a formal and intellectual character […] by deliberate artifice” (Catto 2003: 36) and thus to elevate themselves to the status of English Dantes (Catto 2003: 37). Instead it is interesting to watch what has become of these loans in Modern English, because there are apparently some such loans – like commence – that have never made it into the “immediate” informal register, but rather retain “a distinct stylistic significance even in modern English” (Smith 2006: 128). Although it must be doubted whether any theory of borrowing can be developed to allow valid predictions of what may and what may not enter the informal register, a historical retracing of individual lexical histories
529
530
IV Middle English in terms of discourse traditions would certainly be quite fruitful, at least to account for some such items.
4.2 Syntactic elaboration Similar observations of register-distribution can be made with loans that have contributed to the inventory of free grammatical morphemes with their “[…] emergence and grammaticalisation patterns of new prepositions and adverbial subordinators borrowed from French or Latin and adopted to the gradually developing Southern Standard in the fourteenth and fifteenth centuries” (Rissanen 2006: 133). For the time being, Modern English because, notwithstanding and during may serve as examples. As a partial calque of French a` cause (de), Middle English because was already grammaticalized by the age of Chaucer (Rissanen 2006: 138). In terms of register, because is today clearly favored in conversation, and in terms of frequency as adverbial subordinator it ranks second after if and before when (Biber et al. 1999: 842). In contrast, notwithstanding which translates Fr. nonobstant/Lt. non obstante in legal/administrative texts of the 15th century (cf. Rissanen 2002; Weber 2010) has never lost its “legalese” touch, so that it is today still qualified as “formal and rather legalistic in style” (Quirk et al. 1985: 706). The catchword “legalese” takes us back to the alleged role of “Chancery” – or rather “administrative” – English. Recollecting Wright’s (1994: 113) statement that it is “[…] not adequate to suggest that this diversity of function [of Modern Standard English] could have arisen solely from Chancery documents”, we need to take into account that the number of “government documents” in English after 1430 may only be great when considered in isolation. Unfortunately, the Parliamentary Rolls of Medieval England (PROME), which are now electronically available (Given-Wilson et al. 2005), for various reasons do not allow one to quantify the shares of English, French, and Latin in this corpus. In any event, French and Latin are still high in quantity. What PROME does, however, allow is a look into the side-by-side administrative use of all three languages, and it is this more or less simultaneous use which should be seen as the decisive clue to the extensive and intensive elaboration of written English. In her recent book on the Ausbau of late Middle English, Beatrix Weber (2010) has illustrated how individual elements like Fr./Lt. durant- and its English semicalque during occur in specific collocations in all three languages (Weber 2010: 121– 135 and 178–181), so that it may be surmised that initially Fr./Lt. durant- was not only calqued, but was “transferred” in more or less frequent lexical combinations. Subsequently, the former present participle, grammaticalized into a preposition, could be combined freely, and is now register-neutral, while, as has been said above, the English calque notwithstanding, for instance, remains register-marked. It needs to be repeated that our knowledge of the syntactic elaboration of late Middle English through borrowing and calquing is only slowly unfolding. Nevertheless, we may assume that by conceptualizing the increase of syntactic versatility in late Middle (and early Modern) English as discourse-traditional norm compliance, we will gain further insight into the specific conditions of a rising written standard that shifts from “specific purpose” to an “all-purpose” use. Many of the relevant phenomena will, however, remain restricted to the written “language of distance” (Koch and Oesterreicher 1985),
33 Middle English: Standardization for which they ultimately developed in Latin and French and were then taken over into English.
5 Summary The communicative space of post-Conquest medieval England was trilingual for a long time, and by the late 15th century this situation had not yet come to an end. Nevertheless, the use of English in writing substantially increased in quantity as well as in its functional range. While in the first two centuries of Middle English standardizations are, at best, locally – or even individually – restricted, English texts from the 14th and the 15th centuries show attempts supra-locally to extend the written vernacular’s use by following diatopically neutralized norms of spelling, grammar, and lexis. By reducing dialectal variability, these are standardizing moves toward a “minimal variation in form” (Haugen 1966), although this was not yet achieved by formal codification. While such standardizing developments reduced variation, the extensive elaboration to achieve a “maximal variation in function” demanded increased variation, or rather variability. This type of standardization, and hence norm-compliance, aimed at the discourse-traditional norms already available in the literary languages French and Latin. Late Middle English was thus re-established in writing with the help of both Latin and French norms rather than merely substituting for these literary languages. Last but not least, it has been seen at various points in this article that it does not take physically existing institutions such a school, an administrative center, a poeta laureatus (cf. Horobin, Chapter 36), or even a king to trigger and sanction standardizations. Standardizations rather establish themselves in terms of norm-oriented communicative practices (cf. Nevalainen 2000: 333), and in Middle English such practices reduced variation to serve supra-local purposes, and enhanced variability to serve cross-register versatility.
6 References Benskin, Michael. 1991. Some new perspectives on the origins of standard written English. In: J. A. van Leuvensteijn and Johannes B. Berns (eds.), Dialect and Standard Language: In the English, Dutch, German and Norwegian Language Areas; Seventeen Studies in English or German [proceedings of the Colloquium “Dialect and Standard Language”, Amsterdam, 15–18 October 1990], 71–105. Amsterdam: North-Holland. Benskin, Michael. 2004. Chancery Standard. In: Christian Kay, Carole Hough, and Irene Wotherspoon (eds.), New Perspectives on English Historical Linguistics. Selected Papers from 12 ICEHL, Glasgow, 21–26 August 2002. Vol. II. Lexis and Transmission, 1–40. Amsterdam/ Philadelphia: John Benjamins. Biber, Douglas, Stig Johansson, Geoffrey Leech, Susan Conrad, and Edward Finegan. 1999. Longman Grammar of Spoken and Written English. Harlow: Longman. Black, Merja. 1999. AB or simply A? Reconsidering the case for a standard. Neuphilologische Mitteilungen 100: 155–174. Burnley, John David. 1986. Curial prose in England. Speculum 61: 593–614. Burnley, John David. 1989. Sources of standardization in later Middle English. In: Joseph B. Trahern, Jr. (ed.), Standardizing English: Essays in the History of Language Change. In Honor of John Hurt Fisher, 23–41. Knoxville: The University of Tennessee Press. Burnley, John David. 2001. French and Frenches in fourteenth-century London. In: Dieter Kastovsky and Arthur Mettinger (eds.), Language Contact in the History of English, 17–34. Frankfurt am Main: Peter Lang.
531
532
IV Middle English Catto, Jeremy. 2003. The making of the language 1370–1400. Past & Present 179: 24–59. Chambers, R. W. 1932. On the Continuity of English Prose from Alfred to More and his School. An Extract from the Introduction to Nicholas Harpsfield’s Life of Sir Thomas More edited by E. V. Hitchcock and R. W. Chambers. London: Oxford University Press (for the Early English Text Society). D’Ardenne, S. R. T. O. (ed.). 1961 [1936]. Þe Liflade and te Passiun of Seinte Iuliene. (Early English Text Society 248.) Oxford: Oxford University Press. [First published in the Bibliothe`que de la Faculte´ de Philosophie et Lettres de l’Universite´ de Lie`ge (fsc. LXIV) 1936.] Diller, Hans-Ju¨rgen and Manfred Go¨rlach (eds.). 2001. Towards a History of English as a History of Genres. Heidelberg: Winter. Fisher, John H. 1996 [1977]. Chancery and the emergence of standard written English. In: Fisher, 36–64. [First published in Speculum 52, 1977.] Fisher, John H. 1996 [1992] A language policy for Lancastrian England. In: Fisher, 16–35. [First published in PMLA 107, 1992.] Fisher, John H. 1996. The Emergence of Standard English. Knoxville, KY: The University of Kentucky Press. Given-Wilson, C. P., P. Brand, A. Curry, R. E. Horrox, G. Martin, W. M. Ormrod, and J. R. S. Phillips (eds.). 2005. The Parliament Rolls of Medieval England. Scholarly Digital Editions. http:// www.sd-editions.com/PROME/home.html Go¨rlach, Manfred. 1990a [1988]. The development of Standard Englishes. In: Go¨rlach (ed.), 9–64. Go¨rlach, Manfred. 1990b [1986]. Middle English – a creole? In: Go¨rlach (ed.), 65–78. Go¨rlach, Manfred (ed.). 1990c. Studies in the History of the English Language. Heidelberg: Carl Winter Universita¨ts Verlag. Haugen, Einar. 1966. Dialect, language, nation. American Anthropologist 68: 922–935. Koch, Peter and Wulf Oesterreicher. 1985. Sprache der Na¨he – Sprache der Distanz. Mu¨ndlichkeit und Schriftlichkeit im Spannungsfeld von Sprachtheorie und Sprachgeschichte. Romanisches Jahrbuch 36: 15–43. Koch, Peter and Wulf Oesterreicher. 1994. Schriftlichkeit und Sprache. In: Hartmut Gu¨nther and Otto Ludwig (eds.), Schrift und Schriftlichkeit. Writing and Its Use, 587–604. (HSK 10.1.) Berlin/New York: Mouton de Gruyter. Kurath, Hans, Sherman M. Kuhn, John Reidy, and Robert E. Lewis. 1952–2001. Middle English Dictionary. Ann Arbor: University of Michigan Press. http://quod.lib.umich.edu/m/med/ Liuzza, Roy. 2000. Scribal habit: The evidence of the Old English Gospels. In: Mary Swan and Elaine M. Treharne (eds.), Rewriting Old English in the Twelfth Century, 143–165. Cambridge: Cambridge University Press. Milroy, James. 1992. Middle English dialectology. In: Norman F. Blake (ed.), The Cambridge History of English. Vol. II. 1066–1476, 156–206. Cambridge: Cambridge University Press. Milroy, James. 1996. Linguistic ideology and the Anglo-Saxon lineage of English. In: Juhan Klemola, Merja Kyto¨, and Matti Rissanen (eds.), Speech Past and Present. Studies in English Dialectology in Memory of Ossi Ihalainen, 169–186. Frankfurt am Main: Peter Lang. Nevalainen, Terttu. 2000. Processes of supralocalisation and the rise of Standard English in the Early Modern period. In: Ricardo Bermu´dez-Otero, David Denison, Richard M. Hogg, and C. B. McCully (eds.), Generative Theory and Corpus Studies. A Dialogue from 10 ICEHL, 329–371. Berlin/New York: Mouton de Gruyter. Nevalainen, Terttu. 2004. Letter writing. Introduction. Journal of Historical Pragmatics 5: 181–191. Nevalainen, Terttu. 2006. Fourteenth-century English in a diachronic perspective. In: Schaefer (ed.), 117–132. Ormrod, W. Mark. 2003. The use of English: Language, law, and political culture in fourteenthcentury England. Speculum 78: 750–787. Quirk, Randolph, Sidney Greenbaum, Geoffrey Leech, and Jan Svartvik. 1985. A Comprehensive Grammar of the English Language. London: Longman.
33 Middle English: Standardization
533
Richardson, Malcolm. 1980. Henry V, the English Chancery, and Chancery English. Speculum 55: 726–750. Rissanen, Matti. 2000. Standardisation and the language of early statutes. In: Laura Wright (ed.), The Development of Standard English 1300–1800. Theories, Descriptions, Conflicts, 117–130. Cambridge: Cambridge University Press. Rissanen, Matti. 2002. Despite or notwithstanding? On the development of concessive prepositions in English. In: Andreas Fischer, Gunnel Tottie, and Hans Lehmann (eds.), Text Types and Corpora. Studies in Honour of Udo Fries, 191–203. Tu¨bingen: Gunter Narr. Rissanen, Matti. 2006. On the development of borrowed connectives in fourteenth-century English: Evidence from corpora. In: Schaefer (ed.), 133–146. Samuels, M. L. 1989 [1963]. Some applications of Middle English dialectology. In: Angus McIntosh, M. L. Samuels, and Margaret Laing (ed.), Middle English Dialectology. Essays on Some Principles and Problems, 64–80. Aberdeen: Aberdeen University Press [first published in English Studies 44, 1963]. Sandved, Arthur O. 1981. Prolegomena to a renewed study of the rise of Standard English. In: Michael Benskin and M. L. Samuels (eds.), So Meny People, Longages and Tonges. Philological Essays in Scots and Mediaeval English Presented to Angus McIntosh, 31–42. Edinburgh: Benskin & Samuels. Schaefer, Ursula. 1996. The Late Middle English Paston Letters: A grammatical case in point for reconsidering philological methodologies. In: Ju¨rgen Klein and Dirk Vanderbeke (eds.), Anglistentag 1995 Greifswald, 313–323. Tu¨bingen: Niemeyer Schaefer, Ursula. 2008. “Your Maiestee aue fause Frenche enough” – or: Did Henry V have a language policy? In: Anja Mu¨ller-Wood (ed.), Texting Culture – Culturing Texts. Essays in Honour of Horst Breuer, 43–58. Trier: Wissenschaftlicher Verlag. Schaefer, Ursula (ed.). 2006. The Beginnings of Standardization: Language and Culture in Fourteenth-Century England. Frankfurt am Main: Peter Lang. Smith, Jeremy J. 1992. The use of English. Language contact, dialect variation, and written standardisation during the Middle English period. In: Tim William Machan and Charles T. Scott (eds.), English in Its Social Contexts. Essays in Historical Sociolinguistics, 47–68. New York/ Oxford: Oxford University Press. Smith, Jeremy J. 1996. An Historical Study of English: Function, Form and Change. London: Routledge. Smith, Jeremy J. 2000. Standard language in Early Middle English? In: Irma Taavitsainen, Terttu Nevalainen, Pa¨ivi Pahta, and Matti Rissanen (eds.), Placing Middle English in Context, 125–139. Berlin/New York: Mouton de Gruyter. Smith, Jeremy J. 2006. From Middle English to Early Modern English. In: Linda Mugglestone (ed.), The Oxford History of English, 120–146. Oxford: Oxford University Press. Taavitsainen, Irma. 2001. Language history and the scientific register. In: Diller and Go¨rlach (eds.), 185–202. Thomason, Sarah Grey and Terrence Kaufman. 1988. Language, Contact, Creolization, and Genetic Linguistics. Berkeley: University of California Press. Weber, Beatrix. 2010. Sprachlicher Ausbau: Konzeptionelle Studien zur spa¨tmittelenglischen Schriftsprache. Frankfurt am Main: Peter Lang. Wright, Laura. 1994. On the writing of the history of Standard English. In: Francisco Ferna´ndez, Miguel Fuster, and Juan Jose´ Calvo (eds.), English Historical Linguistics 1992. Papers from the 7th International Conference on English Historical Linguistics. Valencia, 22–26 September 1992, 105–115. Amsterdam/Philadelphia: John Benjamins.
Ursula Schaefer, Dresden (Germany)
534
IV Middle English
34 Middle English: Sociolinguistics 1 2 3 4 5
Introduction Data in Middle English Approaches to Middle English sociolinguistics Summary References
Abstract This chapter describes and discusses the state-of-the-art in historical sociolinguistics, particularly when applied to Middle English. It suggests that all three domains of modern sociolinguistics, i.e. correlational and interactional sociolinguistics and the sociology of language, can be fruitfully employed in the study of Middle English, despite the fact that the Middle English period can indeed be very challenging with respect to language internal and external data. However, contrary to Labov’s claim that historical sociolinguistics is the art of making best use of bad data, it is shown that sociolinguistic studies in Middle English (and other periods) can also provide perspectives which ideally complement findings for Present-day English. In particular, the study of linguistic innovation, actuation, and diffusion across individual speakers’ lifetimes and across generations provides valuable new insights.
1 Introduction The Middle English (ME) period is a “middle” period also with regard to historical sociolinguistics. In some respects the sociolinguistics of Middle English is like modern sociolinguistics, since this appears to be the first period of English when we have enough linguistic and social data that warrant correlational and interactional analysis for at least some groups of society, or some discourse styles. There is even some data about the sociology of language. This is not to say, of course, that these studies are not possible for the time before Middle English. Studies like Toon (1983), Lenker (2000), and Gneuss (1971), among others, have shown that even Old English offers some very interesting viewpoints for historical sociolinguistics. With the ME period, however, we see a significant increase in both sociological and linguistic material. This is also illustrated by the design of the Helsinki Corpus (Rissanen et al. 1991) (see Table 34.1). The ME period has the largest section with more than 600,000 words. However, the Early Modern English period is more balanced in its three subperiods (cf. Kyto¨, Volume 2, Chapter 96). And yet, there is no reason to think that the sociolinguistics of Middle English is anywhere near modern sociolinguistics, as it still lacks a vast amount of data on a number of aspects. So, for example, there are very few texts written by women and the “lower classes” (which, for the ME time is an anachronistic concept, of course, see Bergs 2012). Both these groups of speakers are united in the fact that women and members
Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 534–551
34 Middle English: Sociolinguistics
535
Table 34.1: Data in the Helsinki Corpus (Rissanen et al. 1991) Sub-period OLD ENGLISH I –850 II 850–950 III 950–1050 IV 1050–1150 Total MIDDLE ENGLISH I 1150–1250 II 1250–1350 III 1350–1420 IV 1420–1500 Total EARLY MODERN ENGLISH, BRITISH I 1500–1570 II 1570–1640 III 1640–1710 Total
Words
%
2 92 251 67 413
190 050 630 380 250
0.5 22.3 60.9 16.3 100.0
113 97 184 213 608
010 480 230 850 570
18.6 16.0 30.3 35.1 100.0
190 189 171 551
160 800 040 000
34.5 34.5 31.0 100.0
of the “lower classes” were usually illiterate in the Middle Ages. Graff (1987) estimates that only about 10% of the male population in the late Middle Ages was literate. For the female population we can assume something like 5% literacy in general. However, we also need to keep in mind that these figures varied considerably, depending on the place of living (urban literacy was higher) and social group (cf. Cressy 2006). Nevertheless, it can be argued (see Bergs 2005) that historical sociolinguistics is not “the art of making the best use of bad data” (Labov 1994: 11), but that it offers some very helpful and innovative insights into language variation and language change. One example of these would be the focus on individual speakers and their personal language use over long periods of time. In this chapter, I will first give a brief description of the data that is available in the Middle English period for sociolinguistic analyses. After that, I will present a look into the three major strands of sociolinguistics, i.e. correlational sociolinguistics, interactional sociolinguistics, and sociology of language, and their usability and applicability in the Middle English period. In this part, the aim is to integrate general theoretical and methodological issues with actual exemplary case studies. A final section summarizes the major findings.
2 Data in Middle English Probably one of the most popular sources for Middle English data is the Helsinki Corpus (Rissanen et al. 1991). The number of words in the individual sub-periods, however, already shows that the ME period must not be regarded as very homogeneous. On the contrary, as can be seen from Table 34.1 above, Section I, 1150–1250, contains c.18% of the data, Section II, 1250–1350, about 16%, Section III, 1350–1420, 30% and Section IV, 1420–1500, 35%, i.e. about twice as much as Section II, even though it is 20 years
536
IV Middle English shorter than Section II. Why is that so? We probably have to look into social history in order to explain this uneven distribution. The Norman Conquest took place in 1066. With this military strike the ruling class of Britain literally changed over night and Norman French became the language of many official purposes (cf. Trotter, Volume 2, Chapter 114). Simultaneously, we may assume that the country was still in a great deal of turmoil, even after the conquest. This means that people probably wrote fewer texts in English and that many manuscripts may also have been lost in postwar troubles. From about the middle of the 14th century onwards, we see a revival of the English language. In 1362, for example, The Statute of Pleading stipulated that “all Pleas which shall be pleaded in [any] Courts whatsoever, before any of his Justices whatsoever, or in his other Places, or before any of His other Ministers whatsoever, or in the Courts and Places of any other Lords whatsoever within the Realm, shall be pleaded, shewed, defended, answered, debated, and judged in the English Tongue, and that they be entered and inrolled in Latin” (Statute of Pleading, 1362, 36 Edw. 3, [Stat. 1], ch. 15 [Eng.], cf. Mellinkoff 1963: 111–112) – for the very simple reason that people did not know enough French anymore. So eventually we see a significant increase in our English language data over time between 1100 and 1500. Simultaneously, our data become more diverse, i.e. we find more genres and female authors entering the scene. The first dramatic pieces that we still have today stem from 12th and 13th century France (Querolus by Vitalis de Blois, La Jeu de Robin et Marion by Adam de la Halle), the first English dramatic pieces such as Mummings, Miracle, Morality and Mystery Plays did not appear before the 14th century, it seems. One of the first chivalric romances, imported from France, is the late 13th century Havelok the Dane. Some of the first women writers were Marie de France (12th century), Julian of Norwich (mid 14th century), and Margery Kempe (late 14th century), cf. Arnovick, Chapter 35. One of the most pressing problems for us today, however, is that we need to distinguish between authors and scribes. As has been pointed out before, only very few people in the Middle English period were actually literate. Many people who did not know how to read or write employed scribes (either family members or professional scribes) to write down their thoughts. Sometimes even literate people preferred the services of others. The documents which were not written by the authors themselves may have been created through multiple techniques: they could have been written (a) in real, online dictation, (b) through dictation with asynchronous writing (the scribes first listened to the speakers and then composed the documents from memory), (c) maybe also as products of the scribes themselves who were only provided with keywords and intentions. And, finally, sometimes literate authors would even draft their documents and let professional scribes finish their jobs by finalizing the manuscripts (cf. Clanchy 1993). This would not be a major problem, of course, if we were not looking at this from a sociolinguistic point of view. In sociolinguistics we try to bring social factors into some sort of correlation with linguistic facts. If we cannot be sure who actually produced the linguistic facts, scribes or authors, it is very hard to establish social facts. In some cases, we actually do know the scribes, in other cases we don’t. Either the author or the scribe of a given text can be known, or both, or neither of them. If neither of them is known, the text is completely anonymous. One such text is the 13th century monastic manual
34 Middle English: Sociolinguistics Ancrene Wisse). If both scribe and author are known, we usually know quite a few details about the development of the text or manuscript in question. One example would be Chaucer as the author, and Adam Pinkhurst as the scribe of the Consolation of Philosophy or possibly even the Canterbury Tales. And there are of course cases where we know the author, but not the scribe. William Langland is the author of Piers Plowman, but we do not know who the scribe of Corpus Christi College, MS 201(F) was. We don’t know the author of the Brut, but we know the scribe and translator John Shirley, who composed the copy in Harvard MS English 530 (fol 180v). But then we also need to keep in mind that “knowing” is a relative concept. We know a lot about John Shirley and his life, but only comparatively little about the person Adam Pinkhurst, and even less about the thousands of “Hand A”s and “Hand B”s who riddle our manuscripts. So, even scribal identification does not necessarily mean that we have sufficient sociological data on these people. It goes without saying that the same applies to authors: while we have quite a few Chaucer biographies, there is only little to say about William Langland. Where does that leave us? Back to square one: the art of making the best use of bad data? Not necessarily. Even if we do not know all the details of some author’s or scribe’s life, one can still try to locate these people in their social and cultural environment.
3 Approaches to Middle English sociolinguistics Modern English sociolinguistics can also be useful in the study of Middle English. In the following, each of the three major current approaches (correlational and interactional sociolinguistics as well as sociology of language) will be discussed in an individual section.
3.1 Correlational sociolinguistics Correlational sociolinguistics essentially seeks to uncover any relationship (correlation) between language internal variables – e.g. certain forms of pronunciation, morphological phenomena, or syntactic patterns – and independent language external variables – e.g. gender, class, education, age, or ethnicity. Such a correlation is usually expressed statistically. This in turn means that enough data, both language internally and externally, has to be available in order to reach statistical significance for the specific findings. Correlational sociolinguistics is interesting for historical sociolinguistics, and particularly Middle English, as long as certain social factors and gender are not concerned. We do not have sufficient statistical data and evidence for gendered language use in Middle English. There are a few female authors, such as Margery Kempe and Julian of Norwich (cf. Barratt 2010). Their data, however, are not enough for any statistically significant analysis. It is only towards the end of the ME period, i.e. after about 1400, that we find more linguistic material to warrant comparisons between male and female language use. Similarly, the use of the concept of social class must be seen as an anachronism (see Bergs 2012). Social historians are still not entirely sure when the rise of the class concept in English really took place, but most would agree that medieval society was organized in estates or orders rather than “modern” classes (cf. Horrox and Ormrod 2006). This in turn means that our helpful modern division of lower versus middle
537
538
IV Middle English versus upper class (as in, e.g., Trudgill 1974; Cheshire 1982; Labov 1972, and following) does not apply to Middle English. Rather, one would have to look for corresponding medieval concepts, such as social rank, estate, or order (see Raumolin-Brunberg and Nevalainen 1996, 2003 for an extensive discussion). However, since there is good reason to believe that at least some of the currently relevant factors may also have been operative in some form in the Middle Ages, or even at all times, it would be interesting to investigate them. But even secondary evidence from literature is extremely rare. Chaucer scholars, for example, are still fascinated by the fact that Chaucer used dialect features (i.e. regional, not social variants) in only one of his tales, the Reeve’s Tale, where features of Northern English can be found. Virtually all of his other speakers, no matter what their social background may be, are usually seen as “unmarked” in the linguistic sense. This is also not extremely surprising, considering the fact that standardization was not very strong in the Middle English period. In the 14th and 15th centuries we recognize four different types of incipient standards (see Samuels 1963): Type I: Type II: Type III: Type IV:
Wycliffite language, associated with John Wycliff and his followers, the Lollards. This type mostly represents Central Midlands language. London/Essex language as represented in the Auchinleck manuscript. London language, as exemplified by Chaucer’s language (in the Ellesmere manuscript) Official government language, i.e. the “Chancery Standard”.
Smith (1996: 69) points out that only Types I and IV seem to have spread outside their respective areas and can therefore be regarded as something like modern written standards. However, their status is still nowhere near what we today would call a standard (cf. Smith 1992: 56–57; Schaefer, Chapter 33). This means that in Middle English there was no measuring pole, no objective standard for gauging the language use of individual speakers and speaker groups. There may have been implicit norms of language used in Coseriu’s (1975) sense, but these again usually do not necessarily reflect the conscious evaluation of forms and uses. So we rather need to operate with norms of language use in Coseriu’s sense: what speakers actually and normally do in their linguistic performance can count as the baseline against which the individual’s behavior can be measured. All this does not mean, then, that correlational studies are impossible due to the lack of independent, external data, although Labov (1994: 11) not too long ago claimed that historical sociolinguistics essentially is “the art of making the best use of bad data”. As has been pointed out before, historical sociolinguistics can certainly not compete with presentday approaches, but it can develop interesting and relevant insights. This, however, requires some adjustments and re-keying of certain aspects, such as the use of independent variables. Bergs (2005) for example utilizes the tool of social network analysis to arrive at a correlational analysis of late Middle English morphosyntactic variation. The basic claim is that even though there is not enough representative data from different social groups for the time, there can be enough data from individual speakers. These speakers live and interact within and across various social networks. These network structures are usually measured in density (i.e. their number of ties) and multiplexity (i.e. the complexity of the individual relationships). Network structures may correlate
34 Middle English: Sociolinguistics with the specific linguistic behavior of individuals. Present-day studies such as L. Milroy (1987) and J. Milroy (1992a, b, c) have shown that, on average, dense, multiplex networks correlate with conservative (dialect) language use, while loose-knit, uniplex network structures rather correlate with innovation and change towards the linguistic standard. So, for example, if in a given network most people use -th for the third person singular present, this may be the group norm (NB in a non-prescriptive sense!). A speaker in the network who clearly prefers -s for this function then deviates from the norm/baseline and we may ask if this deviation is facilitated or even made possible by other network structures or a particular lifestyle that this speaker may have. Bergs (2005) investigates three different morphosyntactic variables: plural pronouns (traditional hem/here versus modern them/their), relativizers (the rise of modern wh-structures), and the development of light verb constructions (e.g. give a kiss, have a bath). The database is the 15th century collection of Paston Letters, a corpus of about 240,000 words written between 1421 and 1503 by various members of the Paston family (Davis 1971). For these we have ample social and biographical data available so that a reconstruction of the family network plus network structures that go beyond the family is feasible. The study shows that there is no clear and simple correlation between network structures and language use, but that within comparable groups of speakers, such as the second generation of family members, clear trends can be observed. The more open and uniplex the network structures of a given family member, the more likely this speaker is to use innovative language forms, and the denser and more multiplex their networks, the more conservative speakers tend to be. However, the concrete language use of individual speakers differs intra-individually depending on the variable that is investigated (the more salient and noticeable a form, the less likely it is to change), and on socio-pragmatic contexts. So one complicating factor is that speakers use certain variants not only because of their network structures, but also when they accommodate and dissociate, i.e. when they address a particular person (e.g. conservative language use with superiors) or when they want to achieve certain goals (e.g. conservative language use in letters to parents, asking for money and support). This makes a strict analysis just in terms of network structures more difficult. For that reason, it seems sensible to assume a kind of “variability potential” of individual speakers which depends on their network involvement. A second complicating factor is time. While present-day studies usually focus on one single point in time when analyzing network structures and language use, the Paston study describes and analyzes more than 40 years of language use in one single speaker. During that time we not only see language change in the individual, but also changes in network structures. Figure 34.1 shows the language use of John Paston III between 1461 and 1503. Throughout this time, he has both traditional hem and innovative them available. He begins with a strong preference for the traditional form, but then gradually loses this until he only uses the innovative form after 1479. While the development generally seems to have been gradual, we see a significant drop around the year 1470. It is roughly at that time that his language use seems to have shifted, not only with regard to the pronoun forms, but also in other variables, e.g. the spelling of instead of and instead of . All this can probably lead back to the fact that around the year 1470 we see some events in his life such as a visit to Bruges, the death of his father, and his wooing for Lady Boleyn’s daughter, which may have affected his network structure and social behavior.
539
540
IV Middle English 16 14
hem
12
them
10 8 6 4 2
14 61 14 62 14 64 14 64 14 67 14 68 14 69 14 70 14 71 14 72 14 73 14 74 14 75 14 76 14 77 14 78 14 79 14 80 14 84 14 92 15 03
0
Figure 34.1: John Paston III’s use of hem and them over time
Similar phenomena could also be observed with other speakers, such as John’s mother Margaret, who only began to use innovative language forms after her husband had died in 1466. So, on the one hand we see clear generational shifts between three generations of speakers (see Figures 34.2 and 34.3): generation I is conservative and predominantly uses traditional hem and here almost exclusively; generation II is neither innovative nor conservative and uses both modern and old forms; generation III reverses the situation of generation I and clearly prefers modern them and uses their almost exclusively. On the other hand, John Paston III is a member of the third generation and shows the individual diachronic development which we just described, i.e. he gradually loses the older forms and replaces them with modern ones. In other words, 100% 7 80%
199
272
60%
them hem 40%
22 304
20% 72 0% Generation I
Generation II
Generation III
Figure 34.2: Third person plural pronouns them versus hem in three generations of Pastons
34 Middle English: Sociolinguistics 100%
541
1
80% 111 60% 146 15
their here
40%
20%
66 4
0% Generation I
Generation II
Generation III
Figure 34.3: Third person plural pronouns their versus here in three generations of Pastons
the generational pattern is not clearly reflected in the individual development, and vice versa. All these are complicating factors for network analyses since they require a diachronic “movie” perspective on networks and language variation rather than a “snapshot” one which we usually use in contemporary studies. Still, long-term quantifiable diachronic data also allows us, perhaps for the first time, to look into the linguistic development of individuals across their lifetime. Theoretical questions attached to this include the problem of language change after first language acquisition, and, of course, the interdependence of language internal and external factors. We know only very little about the events that lead to changes in language use in the individual, and even less about how the changing language of the individual speaker ties in with language changes in the community, and thus the language as a whole. So, contrary to what Labov claimed, one could say that, given the right re-keying of questions and methodologies, correlational analyses in historical linguistics, can give us valuable new insights into the language system, its use, and change, even going back as far as the Middle Ages.
3.2 Interactional sociolinguistics There is probably less consensus about how to define interactional sociolinguistics, but I believe it is fair to say that it very broadly deals with “speech as social interaction” (Hudson 1996: Chapter 4). Needless to say, there is no clear-cut boundary between interactional sociolinguistics thus construed, pragmatics, or discourse/conversation analysis. Topics that we commonly find here include language and social identity, i.e. facework, power, and solidarity, male versus female language behavior, taboo, and swearing, to name but a few. These are usually investigated in interactional sociolinguistics not from a quantitative, but rather from a qualitative point of view. So, whereas correlational approaches might ask how often male or female speakers use multiple
542
IV Middle English negation, interactional approaches would rather be interested in how the social role of men and women manifests itself in speech patterns, such as the use of hedges, politeness markers, indirect speech acts, or offensive vocabulary. There is a continually growing body of literature that looks at Middle English from an interactional sociolinguistic point of view. The problem here, however, is that there is hardly any authentic, natural, vernacular data to work with. As a consequence, most studies investigate issues such as swearing, politeness, or verbal dueling in Middle English literature. Bergs (2004), Burnley (2003), and Jucker (2006) offer investigations of the T/V pronouns in (late) Middle English literature, Honegger (2003) looks at forms of address in Chaucer’s The Knight’s Tale, Jucker (2010) examines (im-)politeness in Middle English, Pakkala-Weckstro¨m (2002) offers a study of the language of seduction in three fabliaux. Promises and curses are, among other issues, the topic of Arnovick’s (2006) monograph that aims at tracing orality and spoken performance in Old and Middle English (see Schaefer, Volume Two, Chapter 81). In the following, one exemplary study from the perspective of interactional historical sociolinguistics will be presented. Jucker (2000) presents an analysis of verbal aggression in the late 14th-century Canterbury Tales. He distinguishes between three roles in acts of verbal aggression: speaker, addressee, and target, and four main types of aggression: direct, embedded, mediated, and indirect. Jucker claims that Chaucer “uses a large range of stylistic means to depict verbal aggression. Name-calling, sexual innuendo, scatology, and animal imagery are particularly common” (Jucker 2000: 369). Another important distinction introduced here is that between slanders or slurs and insults. While the former can at least theoretically be tested in terms of truth conditions, the latter are merely about hurting somebody, without any regard to the truth of the utterance. Insults, therefore, can only be identified by the reaction of their target. If the target shows that he or she is hurt or offended, the utterance may have been an insult. Since sometimes reactions are described in literature, the researcher may have to speculate about the possible outcome of such an exchange. One example (1) of an open and direct insult with the appropriate reaction is the following scene, quoted in Jucker (2000: 377): (1)
I rede that oure Hoost here shal bigynne, For he is mosst envoluped in synne. Com forth, sire Hoost, and offre first anon, And thou shalt kisse the relikes everychon, Ye, for a grote! Unbokele anon thy purs. (?1400 Pardoner’s Tale, VI 941–945) ‘I take it that our host here shall begin, For he is most enveloped in sin. Come forth, Sir Host, and offer first at once, And you shall kiss every one of the relics, You, for a groat! Unbuckle your purse at once’ (my translation)
This invitation to kiss the (bogus!) relics provokes the addressee, Harry Bailly, to react most violently and offended. He answers with a counter-insult (2): (2)
“Nay, nay!” quod he, “thane have I Cristes curs! Lat be,” quod he, “it shal nat be, so theech! Thou woldest make me kisse thyn olde breech,
34 Middle English: Sociolinguistics And swere it were a relyk of a seint, Though it were with thy fundement depeint! But, by the croys which that Seint Eleyne fond, I wolde I hadde thy coillons in my hond In stide of relikes or of seintuarie. Lat kutte hem of, I wol thee helpe hem carie; They shul be shryned in an hooges toord!” (?1400 Pardoner’s Tale, VI 946–955; quoted in Jucker 2000: 379–380, 384) ‘ “No, no!” said he, “then I have Christ’s curse! Let be,” said he, “it shall not be, so may I prosper! You want to make me kiss your own old breech, And swear it were a relic of a saint, Though it is soiled with your excrement! But, by the cross which Saint Helen found, I wish I had thy balls in my hand Instead of relics or of sanctuaries. Let’s cut them off, I will help you carry them; They shall be shrined in a hog’s turd!” ’ (my translation) This counter-insult consists not only of a statement (the relics are nothing but old and stained underpants), but also of a highly offensive scatological wish or request. In reaction to this, the Pardoner is simply lost for words (3): (3)
This pardoner answered nat a word; So wroth he was, no word ne wolde he seye (?1400 Pardoner’s Tale VI 956–957; quoted in Jucker 2000: 385) ‘To this, the pardoner answered not a word; So shocked was he, no word did he wish to say’ (my translation)
Jucker (2000: 380) rightly points out that literature is indeed very helpful here. Chaucer not only tells us that the Pardoner falls silent; he also gives the reader the reason why. In real life we would have to include non-verbal behavior, such as noises or facial expressions in our analysis, and would perhaps even have to ask explicitly for reasons. Here, again, historical sociolinguistics has certain advantages over present-day approaches. Insults such as the one given above may be either direct, embedded, or mediated. Embedded insults can be witnessed when one character insults another character in tale. Mediated means that not only one character is the target but also the pilgrims who listen to the tale. Indirect insults are very similar to embedded and mediated insults, but can also occur in the narrative passages. The examples given above also illustrate some of the stylistic means that Jucker identifies in the insults. On the one hand, we have the stained underpants. On the other, Bailly wants to see the Pardoner’s testicles cut off and enshrined in a hog’s turd. Jucker (2000: 384) suggests that scatological references like these still sound very offensive to modern ears. However, many other insults in Middle English are based on expressions like “olde cherl” or “false theef”. These, just like curses “by God” or “by Seint John!”, do not really show the same strength today as they probably
543
544
IV Middle English did in the 14th century. Here, the concrete socio-cultural context of the time must be taken into account. Thus, Taavitsainen (1997) claims that swearing in the Old English period mostly was based on the Anglo-Saxon heroic code, whereas in Middle English religion played a more important role, as does sexuality in the Modern English period. Thus we might suspect that every age and every culture has its specific way of swearing, despite the fact that some general trends can probably be found at all times (religion, sexuality, and bodily functions/fluids being the most common domains of taboo, cf. Allan and Burridge 2006).
3.3 Sociology of language The sociology of language investigates what kind of effect language and language use can have on society, and how far language and language use is shaped by society and societal constraints. Some topics here are multilingualism, code-switching and language choice, standardization and prescriptivism, literacy, language ideologies, and language and power. It is fairly obvious that the time between 1066 and 1476 (or 1100 and 1500, broadly speaking), with all its socio-historical events, offers many interesting and important topics for the perspective of the sociology of language. The most obvious, perhaps, is language choice and multilingualism in the wake of the Norman Conquest (cf. Schendl, Chapter 32). But equally one might also look into the beginning of standardization in the 14th and 15th centuries (cf. Schaefer, Chapter 33), or the loss of the medieval system of the three estates, and the rise of the middle class together with growing urbanization, and their consequences on the language system. How many people and what kind of people did speak what language? What function was associated with what language? And why did English succeed in the end? All these are questions that can and need to be answered from the viewpoint of sociology of language. In the following, I will present three major case studies from this point of view in order to illustrate the approach in general and its potential: language choice (Section 3.1.1), standardization (Section 3.2.2), and the rise of the middle class (Section 3.3.3).
3.3.1 Language choice Let us begin with the question of language choice. After the Norman Conquest, there were at least three languages available in England: English, Latin, and Norman French. After about 1400 one might even subdivide English into two types: English dialects and “standard” English (on the rise of the latter, see Section 3.3.2 below). There is an ongoing debate about which speakers used or at least knew which languages (cf. Trotter, Volume 2, Chapter 114; Skaffari, Volume 2, Chapter 106). From what we know today, it seems likely that most of the French-speaking aristocracy did not speak or know English. And vice versa, most English-speaking commoners probably knew little or no French. The few groups that knew French and English were the interpreters, “latimiers”, more or less high ranking court officials, and some clergymen. This situation seems to have changed over time. Burnley (1992: 424) reports that from the 12th century onwards more and more French speakers tried to learn English and that there is even some rare evidence such as Heloise de
34 Middle English: Sociolinguistics
545
Moreville allegedly calling out to her husband in about 1145: “Huge de Moreville, ware, ware, ware, Lithulf heth his swerd adrage!” (‘Huge de Moreville, beware, Lithulf has his sword drawn!’). But French and Latin certainly remained the languages for serious and technical conversations as well as official documents until about the end of the 13th century. Therefore, the (early) ME period is characterized by something like diglossia with French for “high purposes” and English for “low”. Go¨rlach (1999: 462) develops a very illuminating diagram that summarizes the situation from the earliest days of English to the modern period (Figure 34.4). He analyzes the use of French, Latin, English dialects, and “standard” English in four major varieties of English: law documents, poetry/literature, scholarly texts, and spoken language. Obviously, French and Latin were dominant (but not exclusive) in legal contexts. English dialects and Latin were used alongside a smaller proportion of French in poetry and literature, while scholarly texts were written almost exclusively in Latin and to some degree in French. English dialects were the variety of choice in the spoken channel. Note that for basically all functions, except for spoken language, Standard English took over from about 1400 onwards, reaching exclusiveness in the middle of the 16th century.
English dialects French
Law documents
Standard English Latin
English dialects Poetry, literature
Standard English
French Latin French
Standard English
Scholarly texts Latin
Standard English English dialects
Spoken
French 700
800
900
1000
1100
1200
Latin 1300
1400
1500
1600
1700
Figure 34.4: English dialects, Latin, French and Standard English in four major varieties of English (Go¨rlach 1999: 462)
546
IV Middle English
3.3.2 Standardization The discussion in Section 3.3.1 naturally leads to the question of language development and ideologies. When and how and why did Standard English develop? Numerous studies have looked into this process in great detail (Stein and Tieken-Boon van Ostade [eds.] 1993; Wright [ed.] 2000; Fisher 1995; Schaefer, Chapter 33), so what follows can only be a rough sketch of the first, early phase of this development. Haugen (1972) discusses four steps in standardization processes: selection, codification, acceptance, and elaboration. One particular variety of a given language must be selected as the future standard. This variety needs to be functionally elaborated, i.e. made suitable and available for all communicative functions, for example from informal conversation to religious service and from poetry to textbooks on nuclear physics. Further, standard languages require codification; i.e. they need to have prescriptive grammar books and dictionaries. Finally, the speakers of the language eventually need to accept this new standard as their new standard language. While there is no standard in Middle English in the modern sense, we can still trace at least the beginnings of all of these steps in the Middle English period. The fact that English, rather than French, was chosen as a language for England, does not seem to have been a very conscious decision. It went hand in hand with the elaboration of the functions of English. As Burnley (1992: 428) points out: the grammar teacher John of Cornwall introduced English as the language for the schools in 1349. At the same time, there were attempts at Oxford and Cambridge to “save” the French language. The English parliament officially recognized English alongside French and Latin in 1362 and began to record parliamentary debates in English in 1386. From the early 15th century onwards we witness the use of English in the Chancery, and with it the development of the so-called Chancery Standard in written English. Religious writings in English, including translations of the Bible (another important step in the development of standard languages), can be found from the 14th century onwards (cf. Kohnen, Chapter 65). Conscious efforts at standardizing (written) English can certainly be pinpointed in the 15th century. Caxton introduced the moveable-type printing press in 1476 and discussed varieties of English in his Eneydos, for example. Eventually, he chose the language of London and Westminster as the most widely understood varieties for his books. This, then, might be seen as one of the first steps of choosing a variety of English as the future standard. Codification is another matter. Apart from a few bilingual French-English dictionaries that were produced from the mid-13th century onwards, there are virtually no grammar books or dictionaries before the 16th century. Furthermore, we have hardly any evidence that shows that people thought in categories such as “good” or “bad” language use before that time. What we do find in Chaucer are some comments on regionalisms (in the Reeve’s Tale), but these cannot be understood as comments on proper and improper English. Similarly, we have no evidence about the acceptance of English (and the English of the SouthEast as the variety of choice!) as such. Again, we may assume that this was a gradual process and that discussions of these matters did not become popular before the 16th and 17th centuries (cf. Moessner, Chapter 44).
34 Middle English: Sociolinguistics
3.3.3 The rise of the middle class and urbanization As has been mentioned before, the late Middle Ages and the Early Modern period are times not only of great linguistic, but also of great social changes. The medieval organization of society gradually disappeared and was replaced with our modern class system. At the same time, we see an increase in population and population movement, and the rapid growth of larger cities and urban areas. These changes had some noticeable and interesting impact on the English language, both implicitly (by enabling certain developments) and explicitly (by motivating certain changes). Numerous historical studies, such as Horrox (1994), Horrox and Ormrod (2006), and Britnell (1993), present helpful analyses, and some very helpful summaries can also be found in Nevalainen and Raumolin-Brunberg (1996, 2003). Essentially, what we see is the dissolution of the system described in the anonymous 13th century treatise on the laws and customs of England, entitled Bracton. This text distinguishes people in terms of their being free or not, their having powers over others and/or themselves, their being spiritual or secular leaders, and their being in arms to protect the King and other royalty. The whole system clearly echoes the earlier threefold distinction between warriors, workers, and those who pray. The main point, however, is that there was very little mobility built into the system. Simplifying drastically one could say that those who were born peasants, serfs, or villeins, remained peasants, serfs, or villeins throughout their life without any real hope of leaving this situation by working hard enough. This changed at least from the 13th century onwards. The gradual growth of cities and mercantile operations, together with the ever-increasing strength and power of the guilds and a new “social awareness” (documented, inter alia, by writings such as Piers Plowman or the Lollard treatises, and culminating, maybe, in the Peasants’ Revolt of 1381) ultimately led to the development of a new system with more or less permeable social classes, and, most importantly, a fairly well-to-do and influential middle class. This class is characterized not so much by hereditary rights and noble blood, but rather by money, education, influence, networks, and social aspirations. These people often constitute what has been called the group of “social aspirers”. But why is that important for the language? On the one hand, these social developments – i.e. the population movements to the cities (cf. Keene 2000) and the development of the middle class – enabled (implicitly) the mechanisms of sociolinguistics as we know them today. Social network structures changed and led to supra-regional standardization, the awareness of good and bad language as a marker of social distinction developed, and the middle classes became educated and literate and turned into a powerful driving source of linguistic change (and stability). At the same time, at least one change has been immediately connected with the rise of the middle class: the loss of the ye (you)/thou distinction in the second person singular pronoun. This mostly took place in the Early Modern period, but again we see clear signs of it already in late Middle English (cf. Bergs 2004; Jucker 2006; Mazzon 2000). In the beginning, ye/you and thou were clearly distinguished by their socially determined functions. Ye/you was a plural pronoun and was used among socially high ranking individuals or to address social superiors. Thou was a marker of solidarity among less powerful people and was also used by the high ranks in addressing the low ranks. What we then witness is the pragmaticization of these two forms: instead of only signaling social functions, the two forms may be utilized by speakers when they want to express respect or contempt. When lower rank speakers use you in
547
548
IV Middle English addressing each other (or higher rank towards lower rank) this shows to some degree distance and respect; if a lower rank speaker addresses a higher rank speaker with thou, however, this is usually a sign for contempt and disrespect. Interestingly, the late Middle English Paston Letters do not really show this distinction anymore. In 250,000 words we find only 15 occurrences of thou versus c.2500 of you. Most speakers who should have been addressed with thou receive “polite” you (cf. Bergs 2004, 2005). Note that this situation is different again 150 years later, during Shakespeare’s time. Shakespeare uses the two pronoun forms extensively to signal pragmatic effects (cf. Finkenstaedt 1960; Jucker 2000; Busse 2002). It might be possible that for Shakespeare, these two forms were already quite marked and archaic and thus also available for poetic and dramatic effects and purposes. It has been assumed that the loss of thou is directly linked to the rise of the middle class in the late Middle Ages. With the rise of the middle class, and the blurring of social distinctions, one could not be sure anymore how to address people correctly. It was not only nobility that commanded respect, but also other social ranks, like the new mercantile working class. Even worse, earlier external markers of class membership (like consumption laws) gradually disappeared and left speakers wondering. As early as about 1450 we find evidence for these changes (and complaints about them): It was observed that, in spite of consumption laws, labourers and servants were dressing in more expensive cloth. One preacher of the early fifteenth century was dismayed that a ploughman who would once have been satisfied with a white kirtle and russle gown was now to be seen as proudly dressed as a squire. Peter Idley, writing about 1445–1450, grumbled that ‘a man shall not now ken a knave from a knight’ (Britnell 1993: 169).
Thus, to be on the safe side when talking to each other, people probably opted for the polite form you so that they did not risk being offensive in any way. Eventually, then, the usefulness of thou declined and it gradually went out of use. Thus, there is a direct link between actual changes in social structure and certain linguistic developments. One might even want to speculate about in how far the linguistic situation played a role in the developments (in a feed-and-bleed relationship, maybe, i.e. the social changes led to linguistic changes which in turn may influenced social structures), cf. Busse, Chapter 46.
4 Summary This chapter has described some of the major approaches to Middle English sociolinguistics. In the initial discussion of some general theoretical and methodological issues, it was pointed out that historical sociolinguistics, particularly when it is dealing with fairly early language stages, may need some basic re-keying of established concepts and methods of present-day sociolinguistics. However, it was also claimed that this need not necessarily be a deficit. On the contrary, historical sociolinguistics, particularly in the Middle English period, with its wealth of linguistic variability, can offer new and interesting insights which would be difficult to gain in contemporary studies. In particular, the study of the linguistic individual in society, and that of linguistic developments across long time spans was regarded as most promising. Following this, the paper discussed the three major strands of sociolinguistics, i.e. correlational and interactional sociolinguistics, and the sociology of language in
34 Middle English: Sociolinguistics their particular applicability to and relevance for Middle English. The exemplary case studies show that all three approaches can lead to interesting results and valuable insights, not only from a diachronic, but also from a synchronic perspective. It also became clear, however, that all three approaches have their limitations: the correlational approach certainly needs to cope with the limited availability of language external data, the interactional approach needs to critically discuss the representativeness of data culled from literature, and the sociology of language crucially rests on the availability and reliability of studies from other fields, such as history and historical sociology. Nevertheless, the ME period is still essentially terra incognita for the historical sociolinguistics and future studies will have to explore this domain further.
5 References Allan, Keith and Kate Burridge. 2006. Forbidden Words: Taboo and the Censoring of Language. Cambridge: Cambridge University Press. Arnovick, Leslie K. 2006. Written Reliquaries. Amsterdam/Philadelphia: John Benjamins. Barratt, Alexandra. 2010. Continental women mystics and English readers. In: Carolyn Dinshaw and David Wallace (eds.), The Cambridge Companion to Medieval Women’s Writing, 240–255. Cambridge: Cambridge University Press. ´ lvarez Bergs, Alexander. 2004. Address pronouns in late Middle English. In: Alicia Rodrı´guez-A and Francisco Almeida Alonso (eds.), Voices on the Past. Studies in Old and Middle English Language and Literature, 127–138. Chicago/La Corun˜a: IPG/netbiblio. Bergs, Alexander. 2005. Social Networks and Historical Sociolinguistics. Studies in Morphosyntactic Variation in the Paston Letters (1421–1503). Berlin/New York: Mouton de Gruyter. Bergs, Alexander. 2012. The Uniformitarian Principle and the risk of anachronism in language. In: Juan M. Herna´ndez-Campoy and J. Camilo Conde-Silvestre (eds.), The Blackwell Handbook of Historical Sociolinguistics, 83–101. Oxford: Wiley-Blackwell. Britnell, Richard H. 1993. The Commercialisation of English Society, 1000–1500. Cambridge/New York: Cambridge University Press. Burnley, David. 1992. Lexis and semantics. In: Norman Blake (ed.), The Cambridge History of the English Language. Vol. II. 1066–1476, 409–496. Cambridge: Cambridge University Press. Burnley, David. 2003. The T/V pronouns in later Middle English literature. In: Taavitsainen and Jucker (eds.), 27–46. Busse, Ulrich. 2002. Linguistic Variation in the Shakespeare Corpus. Amsterdam/Philadelphia: John Benjamins. Cheshire, Jenny. 1982. Variation in an English Dialect: A Sociolinguistic Study. Cambridge: Cambridge University Press. Clanchy, Michael T. 1993. From Memory to Written Record, England 1066–1307. Cambridge: Cambridge University Press. Coseriu, Eugenio. 1975. Sprachtheorie und allgemeine Sprachwissenschaft. Mu¨nchen: Fink. Cressy, David. 2006. Literacy and the Social Order: Reading and Writing in Tudor and Stuart England. Cambridge/New York: Cambridge University Press Davis, Norman. 1971. Paston Letters and Papers of the Fifteenth Century. Oxford: Clarendon Press. Finkenstaedt, Thomas. 1960. You and Thou. Studien zur Anrede im Englischen. Berlin/New York: Mouton de Gruyter. Fisher, John H. 1995. The Emergence of Standard English. Lexington: University of Kentucky Press. Gneuss, Helmut. 1971. The origin of Standard Old English and the Æthelwold’s School at Winchester. Anglo-Saxon England 1: 63–83.
549
550
IV Middle English Go¨rlach, Manfred. 1999. Regional and social variation. In: Roger Lass (ed.), The Cambridge History of the English Language. Vol. III. 1476–1776, 459–539. Cambridge: Cambridge University Press. Graff, Harvey J. 1987. The Legacies of Literacy: Continuities and Contradictions in Western Culture and Society. Bloomington, IN: Indiana University Press. Haugen, Einar. 1972. The Ecology of Language: Essay by Einar Haugen. Stanford, CA: Stanford University Press. Honegger, Thomas. 2003. “And if ye wol nat so, my lady sweete, thanne preye I thee, […]”. Forms of address in Chaucer’s The Knight’s Tale. In: Taavitsainen and Jucker (eds.), 61–84. Horrox, Rosemary. 1994. Fifteenth-century Attitudes. Perceptions of Society in Late Medieval England. Cambridge: Cambridge University Press. Horrox, Rosemary and W. Mark Ormrod, 2006. A Social History of England 1200–1500. Cambridge: Cambridge University Press. Hudson, Richard. 1996. Sociolinguistics. Cambridge: Cambridge University Press. Jucker, Andreas H. 2000. Slanders, slurs and insults on the road to Canterbury: Forms of verbal aggression in Chaucer’s Canterbury Tales. In: Irma Taavitsainen, Terttu Nevalainen, Pa¨ivi Pahta, and Matti Rissanen (eds.), Placing Middle English in Context, 369–389. Berlin/ New York: Mouton de Gruyter. Jucker, Andreas H. 2006. “Thou art so loothly and so oold also”: The use of ye and thou in Chaucer’s Canterbury Tales. Anglistik 17(2): 57–72. Jucker, Andreas H. 2010. “In curteisie was set ful muchel hir lest”: Politeness in Middle English. In: Da´niel Ka´da´r and Jonathan Culpeper (eds.), Historical (Im)politeness, 175–200. Bern/ Frankfurt am Main: Peter Lang. Keene, Derek. 2000. Metropolitan values: Migration, mobility and cultural norms, London 1100– 1700. In: Wright (ed.), 93–114. Labov, William. 1972. Sociolinguistic Patterns. Philadelphia: University of Pennsylvania Press. Labov, William. 1994. Principles of Linguistic Change: Internal Factors. Oxford: Blackwell. Lenker, Ursula. 2000. The monasteries of the Benedictine reform and the Winchester school: Model cases of social networks in Anglo-Saxon England. European Journal of English Studies 4: 22–38. Mazzon, Gabriella. 2000. Social relations and forms of address in the Canterbury Tales. In: Dieter Kastovsky and Arthur Mettinger (eds.), The History of English in its Social Context, 135–168. Berlin/New York: Mouton de Gruyter. Mellinkoff, David. 1963. The Language of the Law. London: Pluman. Milroy, James. 1992a. Linguistic Variation and Change: On the Historical Sociolinguistics of English. Oxford: Blackwell. Milroy, James. 1992b. Social network and prestige argument in sociolinguistics. In: Kingsley Bolton and Helen Kwok (eds.), Sociolinguistics Today, 146–162. London: Routledge. Milroy, James. 1992c. A social model for the interpretation of language change. In: Matti Rissanen, Ossi Ihalainen, Terttu Nevalainen, and Irma Taavitsainen (eds.), History of Englishes: New Methods and Interpretations in Historical Linguistics, 72–91. Berlin/New York: Mouton de Gruyter. Milroy, Lesley. 1987. Language and Social Networks. 2nd edn. Oxford: Blackwell. Nevalainen, Terttu and Helena Raumolin-Brunberg (eds.). 1996. Sociolinguistics and Language History: Studies based on the Corpus of Early English Correspondence. Amsterdam/Atlanta GA: Rodopi. Nevalainen, Terrtu and Helena Raumolin-Brunberg. 2003. Historical Sociolinguistics: Language Change in Tudor and Stuart England. London: Longman. Pakkala-Weckstro¨m, Mari. 2002. The discourse of seduction and intrigue: Linguistic strategies in three fabliaux in the Canterbury Tales. Journal of Historical Pragmatics 3: 151–173. Rissanen, Matti, Merja Kyto¨, Leena Kahlas-Tarkka, Matti Kilpio¨, Saara Nevanlinna, Irma Taavitsainen, Terttu Nevalainen, and Helena Raumolin-Brunberg. 1991. The Helsinki Corpus of English Texts. In: ICAME Collection of English Language Corpora (CD-ROM), 2nd edn.,
35 Middle English: Literary language
551
Knut Hofland, Anne Lindebjerg, and Jørn Thunestvedt (eds.), The HIT Centre, University of Bergen, Norway. For manual, see http://khnt.hit.uib.no/icame/manuals/hc/index.htm Raumolin-Brunberg, Helena. 1996. Forms of address in early English correspondence. In: RaumolinBrunberg and Nevalainen (eds.), 167–181. Samuels, Michael L. 1963. Some applications of Middle English dialectology. English Studies 44: 81–94. Smith, Jeremy J. 1992. The use of English: Language contact, dialect variation, and written standardisation during the Middle English period. In: Tim William Machan and Charles T. Scott (eds.), English in its Social Contexts: Essays in Historical Sociolinguistics, 47–68. Oxford/New York: Oxford University Press. Smith, Jeremy J. 1996. An Historical Study of English: Function, Form and Change. New York: Routledge. Stein, Dieter and Ingrid Tieken-Boon van Ostade (eds.). 1993. Towards a Standard English, 1600– 1800. Berlin/New York: Mouton de Gruyter. Taavitsainen, Irma. 1997. Genre conventions: Personal affect in fiction and non-fiction in Early Modern English. In: Matti Rissanen, Merja Kyto¨, and Kirsi Heikkonen (eds.), English in Transition: Corpus-based Studies in Linguistic Variation and Genre Styles, 185–266. Berlin/New York: Mouton de Gruyter. Taavitsainen, Irma and Andreas H. Jucker (eds.). 2003. Diachronic Perspectives on Address Term Systems. Amsterdam/New York: John Benjamins. Toon, Thomas E. 1983. The Politics of Early Old English Sound Change. New York: Academic Press. Trudgill, Peter. 1974. The Social Differentiation of English in Norwich. Cambridge: Cambridge University Press. Wright, Laura (ed.). 2000. The Development of Standard English: Theories, Descriptions, Conflicts. Cambridge: Cambridge University Press.
Alexander Bergs, Osnabru¨ck (Germany)
35 Middle English: Literary language 1 2 3 4 5
Introduction: The notion of Middle English literary language Early Middle English poetry and prose: Alliterative legacies from Old English Middle English texts Summary References
Abstract The notion of literary language proves problematic when it is applied to the Middle English written corpus. While the literary language of Old English did not die entirely with the Conquest, being maintained in a different guise in the alliterative poetry and prose of Middle English, the breach had an inestimable affect on vernacular tradition. Even when the corpus is apportioned by discrete periods of time, a range of textual languages characterize each stage of Middle English. The disparate nature of textual forms and stylistic features becomes apparent in a chronological survey of the period’s
Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 551–576
552
IV Middle English important texts. In the absence of literary language per se, the styles represented in ME writings become the more logical object of study. Select written styles are thus exemplified. This chapter seeks to formulate generalizations about translation as a central form of ME writing and looks at the colloquial language of Corpus Christi plays as representatives of vernacular writing. If a self-consciously literary language does not manifest itself in English until very late in the medieval period, its arrival may be glimpsed in the style of some of the writing reviewed here.
1 Introduction: The notion of Middle English literary language The notion of literary language proves problematic when it is applied to the ME written corpus. The interruption of an English literary tradition by the Norman Conquest is one reason for this. So, too, is a different concept in later medieval culture of what constitutes literature. The great variety of the writings we possess further complicates the idea. Norman Blake locates the difficulty of identifying a specifically literary language in the broad range of texts that comes down to us: “[a]s is true of any period in English, there exists a highly literary style at one end of the spectrum and an equally clear nonliterary style at the other end, but in between there are so many gradations that it is difficult to draw a precise boundary between them” (1992: 500). One of the most striking features of ME literature, taken as a body of work, is, moreover, the linguistic variety it displays. Dialectal differences figure into this impression, as do scribal practices, including the discretion scribes exercised in varying vocabulary and word forms. Medieval writers and scribes were not constrained by the standardization that would occur within the Modern English period, and the literary dialect of Old English was long forgotten. Jeremy Smith likens the variation common in written English of the medieval period to “the degree of variation of the kind now more generally associated with speech” (2009: 149). Synchronic variation in all realms of the grammar is compounded by diachronic changes within a group of texts spanning hundreds of years (roughly 1100–1500). Even when we apportion the corpus by discrete periods of time we find a range of textual languages within each stage of Middle English. The multiplicity of language and textual form resonates within the medieval worldview. Specifically, linguistic and stylistic variety in ME writings seems consistent with contemporary understanding of literature. Unlike our philosophical construct, the post-Conquest notion is more in keeping with the sense of literature’s Latin source, namely littera, ‘that which is written’ (Blake 1992: 500). In other words, the medieval English idea is far more capacious and inclusive than our own. One signal of this is the status of literary documents in the Middle Ages. Manuscripts do not distinguish between these contents. What we now classify as literary texts are not typified by presentations or formats different from those of non-literary texts (Blake 1992: 500). If the language associated with ME literary texts does not necessarily differ from that of non-literary texts, as it does today, it follows that there is a difference in kind between literary language then and now. The phenomenon whereby “literature” identifies itself as such, through marked, self-consciously styled, special language arose only toward the end of the medieval period. Such literary self-consciousness did characterize OE literature, but the line of a continuing tradition was disrupted by the Norman invasion. Correspondingly, we find no contemporaneous discussion of rhetorical strategies that demarcate a literary style from that of any other style of writing. As
35 Middle English: Literary language a result, Blake concludes, no basis exists for formulating a theory of literary language (1992: 502). Few other methods of apprehending literature and literary language have been fruitful so far. Manuscript studies, in particular, the analysis of scribal practices, may someday offer insights into distinguishing forms of language. While scribes may well have treated different kind of texts in particular ways, there is still not sufficient study of the subject. Neither, at this stage in the current scholarship, can we say that certain kinds of glossing practices typify literary, as opposed to non-literary, texts (Blake 1992: 505). For numerous reasons, in summary, description of ME literary language is a thorny pursuit. From our point of departure today, given what we know and what we do not know, we embark on a somewhat questionable undertaking when we seek literature in the first place. If the assumptions underlying our quest are problematic, perhaps it is necessary to redirect our efforts. In the absence of literary language per se, the styles represented in ME writings become the more logical object of study. Even here, we must proceed with caution.
2 Early Middle English poetry and prose: Alliterative legacies from Old English A chronological survey provides a context for the examination of literary styles as they develop over time. With the Conquest, the Germanic tradition of metrical composition gave way to new poetic models; Anglo-Saxon alliterative poetry did not outlive its social and linguistic contexts. The very real demarcation in modes of versification that followed upon the Invasion can be attributed linguistically, “to the clash between the prosodies permitted or required by an outdated poetic canon and the impossibility of such rhythms in contemporary speech” (Minkova 2003: 10). When alliterative verse was next seen in England, it was, therefore, much altered from the earlier form. Two distinct manifestations of poetry characterized by heavy alliteration occurred in medieval England, first in the well-known guise of the 14th-century Alliterative Revival. As we shall see, stylistically these works allude to the Anglo-Saxon tradition without employing its actual techniques of oral composition. Long before this Revival, early in the ME period, however, another, distinct kind of alliterative writing, in poetry and in prose, gained currency. Just as the Alliterative Revival hearkened back to older practices while departing from it, so, too, there was a difference in kind between OE and Early ME alliterative verse. Despite the rupture, the parallels between Anglo-Saxon poetry and alliterative works of the 12th and 13th centuries are striking. While little poetry written in English survives from the period immediately following the Conquest, the earliest extant compositions are alliterative: the Worcester Cathedral fragments of The Soul’s Address to the Body (c.1100, West Midlands), Layamon’s Brut (1189–1205, West Midlands), and The Proverbs of Alfred (c.1175, South-West) (Minkova 2003: 14). Any hypothesis about the connection between OE alliterative verse (see Fulk, this volume) and the alliterative verse of the 12th and 13th centuries is mitigated by a scarcity of evidence and the impossibility of reconstructing the historical circumstances that create and preserve poetry (Minkova 2003: 14). To Minkova, the evidence that remains is highly suggestive of some form of continuity. Analogous traditions of alliterative poetry seem to bridge the divide that separates the literary cultures of Old and Middle English. An
553
554
IV Middle English excerpt from the Brut illustrates Early ME alliterative style. The following passage recounts King Arthur’s dream about the fall of Camelot (1): (1)
Toniht a mine slepe, þer Ich læi on bure, Me imætte a sweuen; þeruore Ich ful sari æm. Me imette þat mon me hof uppen are halle, Þa halle Ich gon bistriden swulc Ich wolde riden. Alle þa lond þa Ich ah, alle Ich þer ouersah; And Walwain sat biuoren me; mi sweord he bar an honde. Þa com Moddred faren þere mid unimete uolke; He bar an his honde ane wi-ax stronge. He bigon to hewene hardliche swiðe, And þa postes forheou alle þa heolden up þa halle. Þer Ich iseh Wenheuer eke, wimmonnen leofuest me; Al þere muche halle rof mid hire honden heo todroh. Þa halle gon to hælden and Ich hæld to grunden, Þat mi riht ærm tobrac. Þa seide Modred: “Haue þat!” Adun ueol þa halle, and Walwain gon to ualle, And feol a þere eorðe; his ærmes breken beine. And Ich igrap mi sweord leofe mid mire leoft honde And smæt of Modred is hafd þat hit wond a þene ueld. And þa quene Ich al tosnaðde mid deore mine sweorede And seoððen Ich heo adun sette in ane swarte putte. And al mi uolc riche sette to fleme, Þat nuste Ich under Criste whar heo bicumen weoren. Buten miseolf Ich gon atstonden uppen ane wolden, And Ich þer wondrien agon wide ʒeond þan moren. Þer Ich isah gripes and grisliche fu ʒeles (Brut ll. 13982–14006; Brook and Leslie [eds.] 1963) ‘Tonight as I was sleeping, where I was lying in my chamber, There came to me a dream which has made me most depressed: I dreamed someone had lifted me right on top of some hall And I was sitting on the hall, astride, as if I was going riding; All the lands which I possess, all of them I was surveying, And Gawain sat in front of me, holding in his hands my sword. Then Modred came marching there with a countless host of men, Carrying in his hand a massive battle-axe. He started to hew, with horrible force, And hacked down all the posts which were holding up the hall. I saw Guinevere there as well, the woman I love best of all: The whole roof of that enormous hall with her hands she was pulling down; The hall started tottering, and I tumbled to the ground, And broke my right arm, at which Modred said ‘Take that!’ Down then fell the hall and Gawain fell as well, Falling on the ground where both his arms were broken, So with my left hand I clutched my beloved sword And struck off Modred’s head and it went rolling over the ground,
35 Middle English: Literary language And I sliced the queen in pieces with my beloved sword, And after that I dropped her into a dingy pit. And all my fine subjects set off in flight, And what in Christendom became of them I had no idea, Except that I was standing by myself in a vast plain, And then I started roaming all around across the moors; There I could see griffins and really gruesome birds’. (Rosamund Allen [trans.]; David and Simpson [eds.] 2006: 125). There is no doubt – as this passage makes clear – that the alliterative poetry of Early ME differs considerably from that of Old English. Its reliance on syllable counting proves only one way in which the later form of alliteration deviates from “classical” Anglo-Saxon rules for versification (Minkova 2003: 16). In Brut, we generally find two stressed syllables in each half-line, both of which are coupled through alliteration, for example (cf. l.13994). As ME poetry developed, certain kinds of verse types with corresponding poetic compounds disappeared. The density of elaborately coined compounds was lost, while kennings became rare (Blake 1992: 509). Resolution is an artifice that does not survive the Conquest (Minkova 2003: 19). These devices are thus absent in Brut, as the sample indicates. Alternatively, new patterns for alliteration emerged (e.g. the tendency to alliterate on initial vowels, vowel- /h- alliteration, f- /v- alliteration); all were innovations based on changes in the contemporary language (Minkova 2003: 19). Assonance in line 13989 of Brut was possible in the new genre, for instance: He bar an his honde ane wi-ax stronge ‘Carrying in his hand a massive battle-axe’. Yet even the discontinuity evident in Middle English can be subsumed into a larger tradition. “After a period of retrenchment alliterative poetry emerged as a vigorous and inspired artistic activity, well adapted to the changed language” (Minkova 2003: 16). Considerable vocabulary originating in Old English was maintained in the Worcester fragments; the French loans found elsewhere in the lexicon had not yet replaced words recorded there by the end of the 12th century (Blake 1992: 509). The fact that the Brut, based on an Old French source, relies on English formulaic phrases traceable to Old English also speaks to the viability of the popular mode (Minkova 2003: 14–15; Blake 1992: 509–512). On the one hand, this poem lacks features prominent in OE verse such as kennings and litotes, as well as the variation of language and the stylistic ornamentation common in the early canon. True apposition occurs once in the twenty-five lines cited above: Wenheuer …, wimmonnen leofuest me ‘Guinevere …., the woman I love best of all’ (l.13992). Major syntactic units also terminate at the end of the line rather than at the caesura, as they do in Old English (Blake 1992: 510): e.g. l.14002, And al mi uolc riche sette to fleme ‘And all my fine subjects set off in flight’. On the other hand, epic formulas, having antecedents in OE poetry, are added to the end of a line of verse to finish the sentence started there without contributing new information (Blake 1992: 510). The second half-line of line 13990 is merely adverbial, for example: Modred hews, ‘with horrible force’ hewene hardliche swiðe. Opportunities for alliteration in the Brut had broadened to include verbs, adverbs, and pronouns, in addition to the main lexical words, nouns and pronouns, on which Old English relied. Alliteration of [s] falls on the adverb, seoððen ‘after’, and the verb sette ‘set’, in line 14001. The two half-lines were usually linked by alliteration, though rhyme was employed as well (Blake 1992: 510). Tobrac ‘broke’ rhymes with þat!, ‘that’, connecting the two half-lines
555
556
IV Middle English of line 13995, at the same time larger cohesion was affected by alliteration throughout the line: þat … þa … þat!. Weighing the evidence en masse, Minkova observes that when poetry in English reemerges, alliteration “persists as an important organizing principle of verse”, despite the new metrical models adopted from the continent (Minkova 2003: 14). A second factor corroborates the notion of poetic continuity, namely, that alliteration continued as a dominating principle in what is called the rhythmical prose of the Early ME period (Minkova 2003: 14). Stylistic features in Layamon’s Brut resemble, if not replicate, features of OE prose works by Ælfric and Wulfstan, as Blake (1969) has established. Blake (1969: 120) labels this form of verse “rhythmical alliteration” to distinguish it from the OE metrical form (cf. Minkova 2003: 14). He hypothesizes that alliterative poetry and alliterative prose are closely related, differing in their rhythmical assumptions, not in kind (Blake 1992: 513). Given that these prose texts survive in copies dating from the 11th and 12th centuries, it is likely that they exerted influence on ME writing. Blake infers a direct line of continuity in which alliterative composition moves from one genre to another and back again: “Classical Old English poetry may have generated Ælfrician prose. In turn, its Middle English successor may have generated the re-creation of alliterative poetry. At first this poetry was of the type found in the Brut, but later in the ME period it developed towards more regular alliteration” (Blake 1992: 513). Alliterative prose thus acts as a graft to reconnect the poetic tradition. Third, despite the limited presence of alliterative poetry in the historical record, we may infer its more widespread composition when the practicalities of manuscript production are taken into account. Minkova reminds us that monastic scriptoria favored the production of Latin and French manuscripts; the copying of English poems was a “special purpose activity” (Pearsall 1977: 90; quoted by Blake 1977: 16). As Minkova (2003: 15) deduces, “[w]hat was interrupted for about a hundred years after the Norman Conquest in England was thus not familiarity with, exposure to, or composition of some form of verse, but the access of English speakers to formal authorial and scribal privileges”. Much of the popular poetry in circulation must have escaped the velum page. Derek Pearsall confronts the lack of material evidence head on, arguing against an inexplicable reversion to the past: “[i]t does not matter that we cannot trace direct lines of descent […] We find instead, at the beginning of the fourteenth century new varieties of alliterative writing […] which bear continued witness to the inherent strength of the alliterative continuum” (Pearsall 1977: 84; quoted by Minkova 2003: 16). Indirectly, then, the nature of the record itself supports a continuity hypothesis. That poetry in English was a popular form and a popular pastime is central to its survival in England. After the Conquest, when a powerful minority who used Latin and French lived alongside a majority of English speakers, the socio-linguistic situation continued to nurture English traditions. A regard for things English, along with its numerous speakers, cannot be dismissed in reconstructing literary styles for Middle English. While it is doubtful that poetry composed before the Conquest, especially in documentary form, was known to new generations of English people, English and its traditions persist (cf. Blake 1992: 508). The deliberate use of English, regardless of subject matter, must not be underestimated. Even if literary preferences changed, the introduction of rhyme and syllable-counting did not overshadow the linguistic potential for an
35 Middle English: Literary language alliterative model of versification, because English is, after all, a stress-timed language (Minkova 2003: 19). A metrical form incorporating Germanic patterns would have been most the accessible choice for the monolingual English-speaking majority (Minkova 2003: 18). Furthermore, as Minkova (2003: 15) argues, the re-invention of alliterative verse in the 14th century is “likely to have drawn upon a lingering perception of alliteration as an appropriate poetic device”. Indeed, by the first half of the 14th century, when written records increased, alliterative verse flourished (Minkova 2003: 18). Minkova (2003: 16) cannot see that this new poetry arose independently of “the inherited appreciation of the mnemonic and artistic power of alliteration”. We may conclude this discussion of 12th- and 13th-century alliterative literature by remarking on one implication of its mixed prosody. Perhaps in its metrical hybridity lies the germ of truly literary language. Early alliterative poetry and prose is marked with a semblance of self-awareness of its power as English. Albeit merely inchoate in the alliterative poetry of Early Middle English, this sensibility would flower into the selfconsciously literary language apparent by the Early Modern period in works such as Spenser’s Faerie Queene.
3 Middle English texts Although aspects of a native English tradition remained, especially in alliterative poetry and prose, the invaders’ French (and Latin) exerted their influence on the genres and styles of English writings throughout the medieval period. A Manual of the Writings of Middle English offers a guide to extant texts and genres (Hartung and Burke Severs [eds.] 1969–93). For the scholarship on these works, see the annotated bibliography published by Burnley and Tajima (1994). Literary traditions from elsewhere in Europe, both classical and vernacular, made their mark, as well. At the height of the Middle Ages, English poets were composing in a kind of rhymed metrical verse borrowed from French. Plot conventions and narrative structure were imported from Italy. Fabliaux and folk stories were transported across the Channel. A vibrant tradition of writing in the ME vernacular coalesced. The range of textual forms and stylistic features found in Middle English becomes apparent in reviewing some of the period’s important texts. Following a chronological overview, I exemplify select written styles. Because the works of major authors like Chaucer and Gower are well known and may be studied in this volume and elsewhere in depth (see Horobin, Chapter 36), this chapter emphasizes the works of their contemporaries in order to illustrate the language of Middle English literature. After this sampling, I attempt to make some generalizations about translation as a central form of ME writing and look at the colloquial language of Corpus Christi plays as representatives of explicitly vernacular writing.
3.1 Early Middle English literature During the 12th and early 13th centuries, fresh copies of OE prose works were made and new prose works were composed. A notable prose piece is the devotional work from the West Midlands, Ancrene Wisse or ‘Guide for Anchoresses’ (c.1225), a monastic rule written for three aristocratic women who leave the secular world to live the contemplative life. Although an extensive mixture of French and Old Norse loanwords
557
558
IV Middle English appear in this text, borrowed vocabulary is used with familiarity. The style is colloquial and sounds spoken, but its alliteration reveals a more poetic or literate register: Hope halt te heorte hal, hwet-se þe flesch drehe; as me seið, ʒef hope nere, heorte tobreke ‘Hope keeps the heart healthy, whatever the flesh may suffer. Without hope, it is said, the heart would break’ (see Hasenfratz [ed.] 2000 from which this quotation is taken; on the language of the manuscripts containing the text, see further Tolkien 1929). The language of Ancrene Wisse is closely related to several other works within the so-called St. Katharine group (also from the West Midlands) and may have shared an intended audience. Early instances of saints’ lives, popular throughout the Middle Ages, appear alongside other didactic works in the St. Katharine group. Stories of the martyrdoms of Saints Margaret, Katherine, and Juliana probably had a general lay audience as well as a religious one, for their narratives appeared in English at a time when Latin and French dominate in this genre (Newhauser 2009: 41). Among early ME prose works we also find the treatise, The Wooing of our Lord, and a series of associated prayers in rhythmic prose (altogether known as the Wooing group, from the West Midlands). Poetry from this early period was both sacred and secular. We have already mentioned The Soul’s Address to the Body, The Proverbs of Alfred, and Layamon’s Brut. From the 12th century we find the Ormulum (c.1170), a work of biblical exegesis, composed in metrical verse by the monk, Orm (or Ormin). Orm’s North-east Midland pronunciation is clearly represented through strict poetic meter and a system of phonetic orthography. Secular lyrics, in the form of the speech contest, Owl and Nightengale (South West), date from the end of the century (1190–1200). The 13th century brings the early romances, King Horn (1225), from the South-east Midlands, Havelok the Dane (c.1280–1300), a North-east Midlands text having a mixture of dialect features, and Sir Orfeo (c.1300), from the South-west Midlands. Highly formulaic and conventionalized, these metrical romances represent popular versions of courtly counterparts. Lacking the literary sophistication common to their French models, these poems reveal a vernacular perception of romantic form and content. In general, several features of Early ME literary style are noticeable in the poetry and the prose of this transitional period. The vocabulary that typified Old English, namely compounds and heroic lexis, is not present, nor is the variety of diction that was common prior to the Conquest (Blake 1992: 514). In the ME works, doublets take the place of the large word hoard and adjectives are introduced to modify nouns. Instead of variation, repetition of the same word throughout the text achieves cohesion. In the passage from Layamon’s Brut cited above, the word, hall, occurs four times in twenty-five lines, while sword, is repeated three times. OE poetry, in contrast, relies on apposition. To complement sword, for example, the poem, Beowulf, employs phrases like ‘best of weapons’ and ‘patterned blade’. Brut similarly lacks Germanic kennings for sword like ‘blood-worm’, ‘icicle of blood’, ‘wound-hoe’, ‘onion of war’, or ‘leek of war’ (all attested in Norse poetry). Words, phrases, and clauses are linked by and (e.g. note nine cases of its use line-initially in the Brut passage) while alliteration often reinforces unity within and between clauses. Stylistically, early texts also depend on the rhetorical conventions of their time (Blake 1992: 515).
35 Middle English: Literary language
3.2 Later ME literature 3.2.1 Prose Religious material flourished from the 14th century onwards, with the circulation of devotional texts written both by clergy and lay theologians. By the end of the medieval period, a Bible in English appeared alongside works offering spiritual direction. Practical guides for public worship attracted English readers, as well. All such publications intended for an English audience marked the prolonged erosion of Latin’s claim to Christian subject matter. Early among instructional works is the moral treatise, Handlyng Synne (1303), an adaptation of a French manual, by Robert Mannyng (North-east Midlands). From c.1300 we find the Cursor Mundi (‘Runner of the World’), an anonymous religious poem written in the Northern dialect. Cursor Mundi subsumes the history of the world into the history of human salvation. Recognizing the value of the vernacular in spreading the message, its author announces that þis ilk bok es translate into Inglis tong/ to rede for the love of Inglis lede,/ Inglis lede of Ingland,/ for the commun at understand (‘This book is translated into the English tongue as advice for the love of English people, English people of England, for all to understand’). Important religious works in prose also survive from the 14th century. The Ayenbite of Inwyt, or ‘Remorse of Conscience’ (1340), is a confessional tract translated or written in ME prose by Michel of Northgate (Kent). Richard Rolle promotes a form of ecstatic mysticism in Form of Living (Northern). He sets out his spiritual advice in 1349 on the occasion of a nun’s moving from the cloister into a solitary cell. Around 1393, the English anchorite, Julian of Norwich, records and interprets her “showings” or visions of Jesus Christ in a book its modern editors often entitle, Revelations of Divine Love (East Midlands), a set of sixteen mystical devotions. Julian’s avowed purpose is to pass along the benefits of her showings to other believers, but the book concerns itself primarily with her own experiences (Blake 1987: 387). Written by a priest whose care is for souls more generally, The Cloud of Unknowing is a mystical treatise dating from 1375–1400 (East Midlands). Borrowing a concept from Dionysian mysticism, the unknown author explains that the contemplative must move beyond the realm of thought into “the cloud of unknowing” (Watson 1999: 552). He also confronts what he believes is the tendency for people who read only the vernacular to understand it literally when the content is meant to be taken spiritually, an interpretation appreciated by the more sophisticated readers of Latin (Watson 1999: 552). A spiritual biography of an ambulatory holy woman, The Book of Margery Kempe (Northern), seems to have been composed by a 15th-century priest with the collaboration of its protagonist. Internal evidence demonstrates that he, rather than the illiterate Margery, actually wrote her “autobiography”. Much of the language of The Book has a clerkly or bureaucratic sound to it (e.g. Whan þe seyde Meyr receyued þe forseyd lettyr ‘when the said Margery received the aforementioned letter’, Book 119: 12–13), a style reinforced overall by its clerical preference for rational chronology (Spearing 2010: 93). Patches of heavy alliteration appear in passages grounded in the tradition of Latinate rhythmic prose, the prosodic affect of which a priest may have tried to imitate (Spearing 2010: 93). While The Book dates from 1436–38 and must have been studied at the Carthusian monastery which owned it as an account of
559
560
IV Middle English affective spirituality, the text in its entirety was not widely known until 1934 (cf. Spearing 2010: 83). Dedication to the life of prayer extends from private devotion to public ritual in another work from this period. The Lay-Folks Mass Book (Northern) is a guide to the Mass for English readers who did not understand Latin (2): (2)
And as þis boke techeth, so þow do, For hit is wretyn what þou schalt say, Whane þow schalt rest, whane þou schalt pray Both for þe quyke and for þe dede; As þow fyndest wryte, so make thy bede … Vpon thi knees sette the doune; And hew vp thyn herte wyt gode entente (The Lay-Folks Mass Book ll. 4–1; Simmons [ed.] 1879). ‘And as this book teaches, so you should do, For it tells you what you shall say, When you shall rest, when you shall pray Both for the living and for the dead; As you find it written, so make your prayer … Upon your knees, set yourself down; And lift up your heart with good intent’.
Originally a translation of a French guide to the Mass, subsequent versions circulated between the 12th and 15th centuries. Prayers in general use in English, like the Lord’s Prayer and bidding prayers, were printed in the Mass Book, but because the Church forbade the translation of sacred liturgy into the vernacular, the book paraphrases Latin prayers for laity untutored in the ecclesiastical tongue. The Mass Book interprets the Eucharistic celebration and directs congregants to perform their own devotions at such times as the priest attends the altar. Reliance on this book by church goers points to increases in literacy in English simultaneous with the loss of Latinity. The Oxford professor, John Wyclif, who advocated teaching the laity in the vernacular, is credited with encouraging the translation of the Latin (Vulgate) Bible into English (see an example below in Section 3.3.2). The ME Bible associated with him comes down to us in the Early Version (finished c.1385) and the Later Version (c.1390) in a dialect of the South-east Midlands (both versions may be accessed online in the Chadwyck-Healey collection, the Bible in English [Hammond and Adamson, eds. 1996–2011]); over two hundred and fifty manuscripts from both textual traditions are extant in a variety of dialects. Both versions were accepted as orthodox at the time, despite the later, common misunderstanding that the Church judged them, as Lollardist productions, heretical (Kelly 2010). Just as it contributed to religious reform, the ME Bible had ramifications for the language, encouraging literacy in English. History and social geography were also popular topics in medieval England. Ralph Higden’s chronicle of world history and theology, Polychronicon, was translated by John of Trevisa in 1387 (West Midlands). In publication well into the 15th century, Caxton prints Trevisa’s version in 1482. One topic of interest was the linguistic situation in Britain at the time: Hit semeþ a greet wonder how Englische, þat is þe burþe tonge of Englisshe men and her owne langage and tonge, is so dyuerse of sown in þis oon ilond
35 Middle English: Literary language ‘It seems a great wonder that English, which is the birth-tongue of Englishmen and their own language and tongue, is so diverse in sound in this one island’ (Babington 1869). Another favorite book of the 14th century was the compendium, Mandeville’s Travels (1360). A translation of a French original from the South-east Midlands, the travelogue of a Sir John Mandeville recounts the places, people, and customs he experiences on a journey throughout the east. Actually a romance, this adventure quest was meant to inspire curiosity in its audience, making it perhaps the most copied and translated book in the Middle Ages (Chism 2009: 68). Romance is the last genre to enter ME prose, having long been the provenance of poetry, as Cooper (2010: 215) notes. Late in the medieval period we find Morte Darthur, or ‘The Death of Arthur’, c.1470, by Sir Thomas Malory (Northern). His English rendition of Arthurian legend was so well received that Caxton printed the romance in 1485. William Caxton himself contributed to the chivalric renaissance. He translated romances originating in Burgundy, namely, Recuyell of the Histories of Troye and History of Jason, as well as others from French, like Blanchardyn and Eglantine, Four Sons of Aymon, and Paris and Vienne (Cooper 2010: 226).
3.2.2 The language of prose translation Because a substantial portion of extant prose derived from the translation of foreign texts, ME translations merit attention in terms of their literary style (on the politics of translating Latin into the vernacular, see further Watson 1999). English versions of Latin and French texts made by Rolle, Trevisa, Malory, and Caxton illuminate translation practices of the time, including those of the Wycliffites. After surveying these strategies, I examine the Bible as a case in point. Ralph Hanna (2010) identifies three kinds of style – “rough,” “middle,” and “high” – in the translations made by Richard Rolle. The meticulous literalism of Rolle’s prose Psalter is “rough” in style. Often employing non-English syntax, this rendition follows the Latin so closely as to invite the English reader to consult the original. Just such a goal motivates Rolle’s decision to seke no strange Inglis ‘seek no unusual English’ for his Psalter, i.e., to avoid any elaborate or compelling language that would detract from the Latin with its clear supremacy in presenting the Psalms. An English Psalter could even act as a liturgical primer: a reader interested in its content and frustrated with unidiomatic expression was thereby encouraged to “come to the Latin” (Hanna 2010: 26). A “middle style” manifests more elegance of diction and phrasing than that found in the rough translations. Hanna exemplifies Rolle’s “middle style” in Form of Living. The passage contemplates the name of Jesus (3): (3)
And whan þou spekes til hym and says “Ihesu” thurgh custom, it sal be in þi ere ioy, in þi mouth hony, and in þi hert melody. For þe sall thynk ioy to here þat name be nevened, swetnes to speke it, myrth and sang to thynk it. If þou thynk Ihesu contynuly and halde it stabely, it purges þi syn and kyndels þi hert; it clarifies þi sawle, it removes anger and dose away slawnes, it woundes in lufe and fulfilles of charite; it chaces þe devel and puttes oute drede; it opens heven and makes a contemplatif man. Have in mynde Jhesu, for al vices and fantomes it puttes owte fra þe lover. And haylce oft Mary, bath day and nyght. Mikel lufe and ioy sal þou fele, if þou
561
562
IV Middle English wil do aftyr þis lare. þe thare noght covayte gretely many bokes; halde lufe in hert and in werke, and þou hase al [done] þat we may say or wryte. For fulnes of þe law es charite; in þat hynges all (108: 4–20; 612–5; Hanna 2010: 27). ‘And when you speak to him and say, “Jesus”, as you are accustomed, it shall be joy in your ear, honey in your mouth, and melody in your heart. For you shall feel joy to hear that name named, sweetness to speak it, mirth and song to think it. If you think “Jesus” continually and hold it firmly, it purges your sin and kindles your heart; it clarifies your soul; it removes anger and does away with sloth. It wounds through love and fulfills by charity; it chases the devil and puts out dread; it opens heaven and makes a contemplative man. Have in mind Jesus: all vices and phantoms it puts away from the lover (of Jesus). And hail often Mary, both day and night. Much love and joy shall you feel if you will follow this teaching. Do not covet numerous books; hold love in your heart and in your work, and you will have done all that we say or write. For charity is the fulfillment of the law. On it all things hinge’. Hanna notes serial presentation and a controlled rush of epithets. Balance is varied in terms of the number of members in each unit, while repetition of detail moves in turn from completeness to incompleteness within these members. Words deriving from foreign and native roots alternate in the vocabulary. If the balance is carefully expansive, the conclusion is nevertheless drawn plainly and directly. The tone grows quieter near the end, members shorten, description is abandoned as the narrator turns directly to the reader, and a generalizing proverb rounds off the passage (Hanna 2010: 27). Rolle’s “high style”, seemingly poised between poetry and prose and between alliterative poetry and septenary verse, has rhythmic cadences and lines embellished with sporadic alliteration and medial and/or end-rhyme (Hanna 2010: 27). Hanna characterizes the high style as a translation of the mannerism of Latin writings. Rolle reserves it for descriptions of the Passion (e.g. “Meditation on the Passion A” and “Ego dormio”) and for incantatory meditations meant to move the reader to ecstasy (Hanna 2010: 27). A preferred technique of John of Trevisa was to translate a single Latin word with two English equivalents (Edwards 2010: 122). In essence he glosses his original, relaying its subtlety of meaning. For example, the doublet, boost and array ‘ostentation and arrangement’ renders the Latin noun pompae, while the pairing, iclosed and ihid ‘closed and concealed’ appears in place of the verb tegebatur in his Polychronicon (Edwards 2010: 122). Edwards (2010: 123) suggests that Trevisa employs doublets when he needs to treat technical material but is at loss to find precise equivalents for terms and concepts unfamiliar to him. Sometimes Trevisa resorts to triplets to define difficult lexemes from Latin. Otherwise his tendency is toward amplification rather than variation. Flexible syntax allows him to maintain control over distinctions in subject matter. Fidelity to his source – and its factual and expository purpose – make his presence as a translator unobtrusive; he identifies himself explicitly when he does intrude, and that is only in signed additions to Higden (Edwards 2010: 124). When we look to prose romances, we see opposing styles in the translations of Caxton and Malory. Caxton was responsible for bringing to England from Philip the Good’s court at Burgundy the new fashion of prosified metrical romances (Cooper 2010: 226). Cooper observes that Caxton himself wrote English as if it were French.
35 Middle English: Literary language In French he found eloquence and ornateness in the form of abundant syntactic subordination, a polysyllabic vocabulary, and rhetorical elaboration (Cooper 2010: 224). He relies on the coupling of synonyms, sometimes pairing an English derivative with one from French (e.g. hookes and crochettes ‘hooks and crooks’). His English prose, whether translation or original, is emotive and expansive. In contrast, Malory displays a minimalist style based firmly in the tradition of writing English (Cooper 2010: 224). He naturalizes French romance with paratactic syntax and English vocabulary. His lexis is “Anglo-Saxon”, with the exception of Anglo-French terms that are primarily legal (e.g. assurance), martial, and chivalric (e.g. joust, recreant), or lie embedded in Arthurian context (e.g. siege), as Cooper (2010: 224) shows. A common tendency in these translations suggests a strategy more widely used: in certain translations, single words in Latin or French are rendered into English with doublets to signal the meanings possible for the foreign word. The resulting translation may not have flowed well in English, but it remained faithful to the original. The move from a literal translation to a more natural English requires further judgment on the part of the translator. The survival of successive biblical translations in Middle English allows us to witness the tie between idiomatic English and this interpretative act. The Early Version (EV; c.1385) of the Wycliffite Bible follows the Latin wording and syntax closely, while the Later Version (LV; c.1390) reflects a more idiomatic English. Revisions made to the earlier Bible admit no kind of amplification; nor does the Later Version depart from the style of the original. Consistently, however, the Late Version avoids the literalness of the Early Version, as we see in Luke 13.8, from the Parable of the Fig Tree (3): (4)
a. Vulgate: Domine, dimitte illam et hoc anno, usque dum fodiam circa illam et mittam stercora. b. EV: Lord, suffre also this yeer, til the while I delve aboute it, and sende toordis. [‘Lord, leave (it) also this year, until I dig around it and spread manure’.] c. LV: Lord, suffre it also this yeer, the while I delve aboute it, and I schal donge it. [‘Lord, leave (it) also this year, while I dig around it, and I shall spread manure’.]
The purpose of the Early Version is to convey the Latin construction, to make the Latin recognizable in English. Viable alternatives in meaning are presented in doublets linked with “either” or “or”, as we see in Luke 6.41, example (5): (5)
a. Vulgate: Quid autem vides festucam in oculo fratris tui b. EV Sothly what seest thou in thy brotheris iye a festu, othir a mot [‘Truly, (why do) you see in your brother’s eye a rod or a mote?’] c. LV And what seest thou in thy brotheris iye a moot [‘(Why do) you see in your brother’s eye a mote?’]
563
564
IV Middle English In this way, the Early Version is consistent with, or may have even been inspired by, the kind of rough translation Rolle produced of the Psalter (cf. Hanna 2010: 26). Where the Early Version offers alternative meanings, the later translation offers a single form as an English equivalent. The later translator, who calls himself “Simple Creature”, makes a choice between possible meanings when these exist or proffers his own substitutions when the Latin is ambiguous. He eliminates ablative absolute constructions, for example, which are retained consistently in the Early Version. In doing so, Simple Creature erases potential ambiguities found in the Vulgate Bible. The Bible remains a special case in translation practice because shades of meaning, even ambiguity, might be thought desirable in a book considered sacred. Both versions of the ME Bible would have held value for students at the time, nonetheless. The Early Version is intended for people who know Latin, the Late Version for people who do not. Blake (1992: 528) remarks that translations which closely follow the original text seem to have been “relatively unself-conscious” until the mid-14th century. Perhaps to the extent that the Late Version of the Wycliffite Bible departs from the Early Version in order to improve its readability in English, the translation approaches self-consciousness. Certainly the formal content of the Bible requires a suitable medium of conveyance, whether Latin or English. In the Wycliffite translation we witness Middle English standing alone as a language worthy of Scripture. To wrap up this discussion of translation practices, let me turn briefly to an analogue in Chaucer’s Canterbury Tales. Malory’s minimalist style and scarce use of elaboration can be compared with Chaucer’s Tale of Melibee. Out of 923 lines (ll. 667–1889), there is a passage of thirty-five lines, ll. 1612–47, in which Chaucer introduces doublets. (This and subsequent references to Chaucer’s works are taken from The Riverside Chaucer edited by Benson [1987].) In line 1612, for example, Chaucer translates purquoy initially as wherto and why and later as for what cause or enchesoun. Doublets appear suddenly in the context of the rest of the tale, according to Kelly (p.c., 2010). Of these twenty pairings, fourteen are coupled by and, five by or, and one by ne. Among twenty doublets occurring within these thirty-five lines, only three of them are found in Chaucer’s French source (Renaud de Louens). Evidence from poetry therefore suggests that the practice of using doublets to translate single foreign forms may be universalized across genres.
3.2.3 Alliterative poetry Poems belonging to the Late ME alliterative canon include Wynnere and Wastoure (1352–1353), from the West Midlands, and Parliament of the Three Ages, a poem difficult to date having Northern dialect features. The manuscript, Cotton Nero A.x (1375– 1400) contains four alliterative works: the romance, Sir Gawain and the Green Knight, and the dream vision, Pearl, along with two biblical narratives, Patience and Cleanness. All are written in the North-west Midlands dialect of Middle English by an unknown author. Another important work in this style, the allegorical poem, Piers Plowman, attributed to William Langland, survives in three versions of different lengths, the A (1365–1370), the B (1376–1379), and the C (1381–1390). This group of poems has been said to represent, and therefore to belong to, an Alliterative Revival. Although the model of revival may not best describe the rise of the
35 Middle English: Literary language 14th-century phenomenon, the nomenclature retains some currency. A note about the construct is, therefore, in order. Hanna (1999) finds the term “revival” reductive. To his way of thinking, the assumption behind this classification was that alliterative usage “had died and at some later point experienced a quasi-divine resuscitation, and that this return to life comprised a single ‘revival’ ” (Hanna 1999: 488). Minkova (2003: 19) prefers the term “renewal” rather than “revival” because it corresponds to “the natural way in which languages and cultural traditions continuously replenish themselves at any time”. Alliterative renewal in the 14th century seems all the more plausible in light of the linguistic evidence for earlier ME. It is likely, as we have seen, that some sort of alliterative tradition was inherited from the Anglo-Saxons, was adapted, and was then sustained through the early ME period. Later ME poetry departed even further from the conventions governing OE composition. Certainly the newer poetry affected older styles, but its underlying structures differed from the ancient patterns. Characteristic of the alliterative poetry of the 14th and 15th centuries, for example, was a “common assumption about vocabulary and content which lies behind all the poems” (Blake 1992: 521). Typically the meter demanded at least three alliterating words per line, a requirement that called for an expansive word hoard. As a result, this group of poems contain a vocabulary distinct from the rest of Middle English. This specialized vocabulary employs forms that are archaic elsewhere in the language as well as large inventories of synonyms for basic vocabulary words like man, e.g. beryn, douth, freke, gome, hathelle, lede, segge, and wy (Blake 1992: 522). Dialectal forms are appropriated for the same reason. Whereas OE alliterative poetry met its lexical needs through a large poetic vocabulary that included compounds, through kennings, and through formulas, Middle English relied on collocations of two or more words that can act similarly in the verse, e,g. foule in that frythe ‘bird in the forest’ and hertys and hyndes ‘male and female red deer’ (Blake 1992: 521–522). Similarly, verbs, adverbs, and pronouns alliterate in Middle English where Old English relies primarily on nominals. Nor is alliteration the sole metrical force in ME alliterative poetry, though it dominates in Old English. Pearl, for instance, is heavily alliterative, appearing in twelve-line stanzas which rhyme ababababbcbc. In the seven lines quoted below, Pearl’s father accepts Christ’s will both for his daughter and for himself (6): (6)
“O Perle,” quoþ I, “of rych renoun, So watz hit me dere þat þou can deme In þys veray avysyoun! If hit be ueray and soth sermoun Þat þou so strykez in garlande gay, So wel is me in þys doel-doungoun Þat þou art to þat Prynsez paye.” (Pearl ll. 1182–1188, Andrew and Waldron [eds.] 2004) ‘ “Oh Pearl”, I said, “of splendid renown, It was so dear to me what you related In your truthful vision; If it be an accurate and truthful speech, That you made in your radiant garland,
565
566
IV Middle English So I am well in this sorrowful fortress That you are pleasing to that Prince” ’. Similarly, a poem like Sir Gawain and the Green Knight constitutes “a mixture of latterday Germanic verse”, unrhymed verse, and a five-line rhyming “bob and wheel” (Renoir 1988: 176). Finally we must remember that OE alliterative poetry was oral in nature, while the ME corpus was thoroughly written. Whereas the Anglo-Saxon poems represent either records of oral poems or written poems composed through the use of oral-traditional devices, ME verse is not, at its basis, oral, even if the older poetic tradition is evoked in its texts. Yet the poetry deploys structures that seem to hearken back to the oralpoetic devices used in OE poetic composition. While these may no longer be productive in the act of generating the poem, or even truly affective outside of their performative context, they lend an archaic legitimacy to the text being read or recited, to its performance as poetry (cf. Renoir 1988: 169). They provide a veneer of indigenous Englishness. See further Thorlac Turville-Petre (1996) on literature, language, and the construction of national identity. The very literate author of Sir Gawain and the Green Knight alludes to oral composition. I schal telle hit as-tit, as I in toun herde,/ with tonge, ‘I shall tell it quickly as I heard [it told] in town with tongue’, the narrator vows, for instance, remembering an oral story-telling tradition, as much as conventional authority, in setting the stage for his romance (Arnovick 2006: 21; Tolkien and Gordon 1967: 31–32; Renoir 1988). Type-scenes are introduced, like that of dressing and arming of the hero before battle, a scene we find enacted before Gawain’s quest for the Green Knight. Sir Gawain’s arrival at Bertilak’s castle may also echo the “hero on the beach” theme found in oral poems like Beowulf. The castle schemered & schon þurʒ þe schyre oke, ‘shimmered and shone through the shining oaks’, just as rays of light typically reflect on the hero’s armor when, disembarking, he arrives at his destination (l. 772; Renoir 1988: 170). Likewise, the moat stands in for the beach and the knights and squires who greet Gawain for the hero’s retainers (Renoir 1988: 170). Without a doubt, as Benson has argued, the poem is “deeply indebted to the tradition of oral verse” (1965: 130; Renoir 1988: 170). In the absence of an active oral tradition, however, we discover “written poetry adorned with an inactive veneer of oral-formulaic rhetoric” (Renoir 1988: 173). Altogether the rhetoric represented a self-conscious allusion to an art by then lost.
3.2.4 Non-alliterative poetry The three great writers of late Middle English literature chose rhymed isosyllabic meter rather than alliteration as the phonological glue for their verse. After a necessarily toobrief mention of Gower and Lydgate’s works, we look closely at Chaucer for his contribution to the development of non-alliterative poetry in English. John Gower’s long narrative poem, Confessio Amantis (‘The Lover’s Confession’) from 1390–1393, must be mentioned first. Although he also composes poetry in Latin and Anglo-French, Confessio Amantis is written in English, in the dialect of London (with Kentish features). In the following passage we observe Gower’s typical octosyllabic line – the standard English meter before Chaucer – and rhyming couplets (7):
35 Middle English: Literary language (7)
And all þis made avant of pride. Good is þerefore a man to hide His oghne pris, for if he speke He may lihtliche his þonk tobreke. In armes liþ non avantance To him which þenkþ his name avance And be renomed of his dede; And also who þat þenkþ to spede Of love he mai him noght avaunte; For what man þilke vice haunte His pourpos schal ful ofte faile. In armes he þat wol travaile, Or elles loves grace atteigne, His lose tunge he mot restreigne, Which berþ of his honour þe keie (Confessio Amantis, Book I, ll. 2647–2661; Macaulay 1900–01) ‘And all this made boast of pride. Good it is therefore for a man to conceal His own reputation, for if he speak He may easily break his reward in pieces. In deeds of arms no boasting is appropriate To him who wants to advance his name And be renown for his deed; And also who wants to prosper Of love he may boast to no avail For whoever practices the same imperfection His purpose shall very often fail In deeds of arms he will labor, Or else love’s favor attain His loose tongue he must restrain Which bears of his honor the key’.
Romance loans words are numerous (e.g. avantance, haunte, travaile, grace, atteigne, honour), revealing Gower’s mastery of Latin and French as much as the borrowed vocabulary of Late Middle English. Gower describes the style as plain, perhaps in keeping with his decision to use the vernacular rather than French or Latin, traditionally (the elite) languages of poetry. In other poets’ hands, a Latinate vocabulary may serve a more formal or elaborate style. The monk and poet, John Lydgate, was much beloved in the 15th century for his Troy Book (1412–1420), Siege of Thebes (1421–1422), Fall of Princes (1431–1438), and many other works (South-west Midlands). Lydgate coins the term auryat, ‘gilded’, to describe his ideal of an elevated, often ornate, poetic diction, often Latinate in origin, suitable for the genre and its subject matter: e.g. adventure ‘adventure’, perfect ‘perfect’ (Blake 1992: 528). Lydgate seems to have modeled his language on the emerging London standard, although his poetry shares characteristics with Chaucer’s particular variety of Middle English (Horobin 2003: 130, 137).
567
568
IV Middle English Although Chaucer brought French and Italian literary traditions to England, ultimately he demonstrated the literary capacity of the English vernacular as a poetic medium. He writes in the dialect of London. Chaucer’s works include the Book of the Duchess (c.1360), The Romaunt of the Rose (1360–1370), House of Fame (1380), Parliament of Fowls (1382), Legend of Good Women (1385), Troilus and Criseyede (1381–1385), and The Canterbury Tales (1386–1400) (see Horobin, Chapter 36; see also Burnley 1983, Cannon 1998, and 2008. Horobin 2003 compares and contrasts the orthography of the various Chaucer manuscripts, isolating Chaucer’s usage from that of his scribes). Chaucer’s gift to English poetry is the rhymed couplet using iambic pentameter. Rhymed iambic verse became the governing model for centuries following its innovation in the 14th century. To consider this verse form in detail, we must address first its larger context, the prosodic features of non-alliterative poetry in Middle English more generally. Through contact with the Anglo-Normans, a new type of versification, one based on syllable counting instead of alliteration, spread across England. A rhymed isosyllabic iambic line became popular (Minkova 2007: 183). Isosyllabism, the great innovation of ME poetry, is based on the iteration of equal measures or feet across verse lines. It is important to stress that the “long” isosyllabic line of ten (or eleven) syllables is not unique to Chaucer. Eight (or nine) syllable lines (couplets or larger groups linked by end rhyme) were the rule in medieval England until Chaucer began to use ten syllables (or five feet per line) in his later poetry. Chaucer borrowed the verse form and anglicized the line by restructuring it rhythmically. Whereas the French line peaked on both the fourth and tenth syllable, for example, Chaucer combined the ten-syllable count with stress alternation: five iambs alternate weak and strong stress (Minkova 2007: 183). Note that inflectional endings containing a vowel are treated as syllables except when final vowels are elided before another vowel and before h when silent or in an unstressed syllable (see further Burrow and Turville-Petre 2005: 5). Line 17 of the General Prologue, The hooly blisful martir for to seke ‘The holy blissful martyr to seek’ illustrates iambic pentameter at its most basic (or unproblematic) operation (Minkova 2007: 190): The hoo ly blis ful mar tir for to se ke W–S W–S W–S W–S W–S (W)
Usually, as it does above in line 17 of the General Prologue, a light syllable precedes each of the stressed syllables in the line (Davis 1987: xliv), example (8): (8)
Bifil that in that seson on a day, In Southwerk at the Tabard as I lay (General Prologue ll. 19–20; Benson [ed.] 1987) ‘It happened in that season one day In Southwark at the Tabard as I lay’
But, as Norman Davis (1987: xliv) explains, this order of stressed and light syllables is also reversed at times, while even headless (i.e. there is no initial light syllable) lines occur in Chaucer’s poetry. Then, too, more light syllables than are required in regular rhythm sometimes occur. The result is “great freedom and a variety of movement”,
35 Middle English: Literary language when Chaucer alters the rhythm by these and other means in both his five- and fourstress lines (Davis 1987: xliv). (See further Minkova 2007: 188–191 on the rhythmic adjustments Chaucer makes to his iambic pentameter.) Returning to Chaucer’s major poetry, we may now classify his versification. Chaucer rhymes couplets of iambic fivesyllable lines in The Legend of Good Women and most of The Canterbury Tales. He also employs the same kind of line in a seven-line stanza (known later as “rime royal”) which rhymes ababbcc in Parliament of Fowls, Troilus and Criseyde, several of the tales told in The Canterbury Tales (i.e. Man of Law, Clerk, Prioress, Second Nun), and in other poems.
3.2.5 Drama As we have seen, the style that medieval writers adopted determined the language they subsequently employed. Richard Rolle matches the style of his translations to their larger purpose, as do the translators of the ME Bible. Chaucer is adept at fitting the language of his narrators to their literary personalities. Chaucer’s Canterbury pilgrims (think of the Wife of Bath) delight us for the plain and often earthy language that comes out of their mouths. As a subject of the Trivium, rhetoric was studied and practiced by medieval writers, although the most skillful employ it with great subtlety. Medieval notions of literary style rooted themselves in the discipline of rhetoric. Medieval rhetoricians taught, for example, that letters should be written to their recipients on the basis of their contents and according to the status of their addressee. The ars dictaminis, a manual of epistolary technique, recommended three appropriate styles: high, middle, and low (Ginsberg 1987: 879). When the Host demands that the Clerk use a plain style rather than a learned style in telling his tale, he applies this notion to oral narrative (9): (9)
Telle us som murie thyng of aventures. Youre termes, youre colours, and your figures, Keepe hem in stoor til so be ye endite Heigh style, as when that men to kynges write. Speketh so pleyn at this tyme, we yow preye, That we may understonde what ye seye (Clerk’s Prologue ll. 15–20; Benson [ed.] 1987). ‘Tell us some cheerful tale of adventures, Your terminology, your colors (i.e. figures of speech), and your figures (i.e. patterns of thought), Keep them in store until you compose High style, as when men write to kings. Speak so plainly at this time, we pray you, That we can understand what you say’
From the Host’s request we realize that Chaucer recognizes two levels of literary style: high and low (Horobin 2007: 126). A low style was one that everyone could understand, as the Host stipulates, whereas a high style was more appropriate for an elite audience. In contrast with low style, which remains unmarked, high style was marked linguistically. The language of high style is distinguished by a restricted vocabulary and
569
570
IV Middle English rhetorical figures of speech (e.g. metaphor, synecdoche). These figures included unusual structures or organizations (e.g. anaphora) as well as sound patterns (e.g. alliteration), devices exploited to achieve a particular rhetorical effect. High style constituted an elevated form for “enditing”, “composing in writing”, requiring, in other words, a heightened linguistic register that one might use for special situations, like addressing a king (Horobin 2007: 127). The Clerk defers to the Host’s preference. Indeed, the stories told by Canterbury pilgrims often adopt the low style heard in common speech. Low style is no less artistically employed than its counterpart high style by medieval writers. Nowhere is this more apparent than in the dialogue written for the Mystery plays, especially by the so-called Wakefield Master. Although medieval drama is too large of a topic for this essay on ME literary language, the Wakefield plays in the Towneley cycle represent late ME literature at its best. For this reason their language deserves further discussion. Produced by local craft guilds to celebrate the Feast of Corpus Christi, several cycles of “Mystery” pageants come down to us, including those of York (Northern, c.1467), Towneley (Northern, 1450–1500), Chester (North-west Midlands, 1591–1607), and N-town (1468–1500). Within the Towneley cycle, the set of plays ascribed to the Wakefield Master possess a poetic style worthy of elaboration here. These plays are distinguished by the form of their compressed nine-line stanza, interpreted in the edition of Stevens and Cawley (1994) as thirteen lines which follow the rhyme scheme, ababababcddc. See further Stevens and Cawley (1994, Vol. 1: xxix–xxxii) on the manuscript and their rationale for printing the Wakefield stanza with thirteen-lines. A colloquial style of dialogue (low style) supports the playwright’s extensive use of anachronism as a literary device. These Corpus Christi plays reenact episodes from the history of the Bible, making relevant its ancient narratives by setting them in medieval time and English place. See Kolve (1966) for an in-depth analysis of the Corpus Christi Plays and their setting. Prosody effects phonological and lexical cohesion in the Wakefield plays. “Secunda Pastorum”, or the “Second Shepherd’s Play”, tells the story of the Nativity. Note that the practice of giving Latin titles to the plays is an old editorial convention; I include them here for readers familiar with that identification. It opens with a 14th-century shepherd lamenting both his state and the cold (10): (10) Lord, what these weders ar cold! And I am yll happyd; I am nerehande dold, So long haue I nappyd; My legys thay fold, My fyngers ar chappyd. It is not as I wold, For I am al lappyd In sorrow. (“Secunda Pastorum”, ll. 1–9. This and all subsequent references to the Towneley plays are taken from Stevens and Cawley [eds.] 1994). ‘Lord, that these weathers are cold! And I am poorly clothed; I am nearly numb, So long have I napped; My legs, they cramp,
35 Middle English: Literary language My fingers are chapped. It is not as I would wish it, For I am wrapped, In sorrow’. The need for rhyme words presses a varied vocabulary into service, as we see in the first stanza: dold ‘stupid’, for example, is used to rhyme with cold and fold in the a verses, while happyd ‘covered’, nappyd ‘napped’, chappyd ‘chapped’, lappyd ‘wrapped’ rhyme in the b verses. In the stanza immediately following, hamyd ‘hamstrung’, ramyd ‘oppressed’, and handtamyd ‘submissive’ appear in the d verses (“Secunda Pastorum,” ll. 14–26). Alliteration is another prosodic feature the Master employs. It appears most noticeably at the religious and dramatic nexus of “Secunda Pastorum”. The First Shepherd recognizes God himself in the newborn babe (11): (11) Hayll, comly and clene! Hayll, yong child! Hayll, maker, as I meyne, Of a madyn so mylde! Thou has waryd, I weyne, the warlo so wylde (“Secunda Pastorum”, ll. 1024–1029) ‘Hail, comely and clean! Hail, young child! Hail, Maker (as I mean), Born of a maiden so mild! You have cursed, I believe, the warlock so wild [i.e. the devil]’. In parallel fashion, both the Second (12) and Third (13) shepherds acknowledge the Incarnation: (12) 2 Pastor. Hayll, sufferan sauyoure, For thou has vs sought! Hayll, frely foyde and flour, That all thyng has wroght! Hayll, full of fauour, That made all of noght! (“Secunda Pastorum”, ll. 1037–1042) ‘Second Shepherd. Hail sovereign savior, For you have us sought! Hail, noble child and flower, That all things has wrought! Hail, full of favor, That made all of nought!’ (13) 3 Pastor. Hayll, derlyng dere, Full of Godhede! I pray the be nere
571
572
IV Middle English When that I haue nede. Hayll, swete is thy chere! My hart would blede (“Secunda Pastorum”, ll. 1050–1055) ‘Third Shepherd. Hail, darling dear, Full of godhead! I pray you be near When I have need. Hail, sweet is your cheer! My heart would bleed’ The alliteration is pronounced and proves functional in this dialogue. The metrical device connects each of the shepherd’s speeches, demarcating them aurally from the rest of the discourse, in which alliteration is less heavy. Alliteration unifies the discourse on a structural level, as well: the consonants found in the last line of one speech begin the first line of the next speech (14): (14) a. (1 Pastor.) Haue a bob of cherys … (l. 1036) ‘(First Shepherd.) Have a bunch of cherries’ b. 2 Pastor. Hayll, sufferan sauyoure … (l. 1037) ‘Second Shepherd. Hail, sovereign savior’ c. (2 Pastor.) I wold drynk on thy cop, Lytyll day-starne (1l. 1048–1049) ‘(Second Shepherd.) I would drink from your cup, Little day-star!’ d. 3 Pastor. Hayll, derlyng dere, (l. 1050) ‘Third Shepherd. Hail, darling dear’ Most important, alliteration signals the climax of “Secunda Pastorum” at the Nativity, for the heavily alliterative style elevates the shepherd’s address. Prosody marks the adoration they offer. Alongside the formal constraints of the prosody is a colloquial style of expression which renders sacred subject matter accessible to a lay, often unlettered, audience. Actual-sounding, informal speech makes the subject relevant on a personal level. “I don’t care a pin” (Set I not at a pyn), Noah’s wife, Uxor, responds to daughters-inlaw insistent that she board his ark (“Processus Noe cum Filiis”, l. 527). Just as contemporary utterances personalize the narrative, colloquial idiom also inevitably creates anachronism. Anachronism allows the remoteness of the past, the time of biblical record, to collapse into the immediacy of the medieval here and now. In turn, both past and present subsume themselves in turn into God’s universal time. Filled with the resentment of any peasant, the First Shepherd complains that he works so that the gentlery-men ‘gentry’ (“Secunda Pastorum”, l. 26) may thrive (15): (15) Bot we sely husbandys That walkys on the moore, In fayth we ar nerehandys Outt of the doore. No wonder, as it standys,
35 Middle English: Literary language If we be poore, For the tylthe of oure landys Lyys falow as the floore, As ye ken. We are so hamyd, For taxed and ramyd, We ar mayde handtamyd With thyse gentlery-men (“Secunda Pastorum”, ll. 14–26) ‘But we miserable husbandmen That walk on the moor, In faith, we are nearly Out of the door [i.e. destitute]. No wonder, as it stands, If we are poor, For our arable land Lies as fallow as the low-lying land, As you know. We are so hamstrung, Overburdened and oppressed, We are made submissive, By these gentry men’. The setting is established with a wink and a nod when the First Shepherd moans, on waking from sleep (16): (16)
And I water fastand I thoght that we layd vs Full nere Yngland (“Secunda Pastorum”, ll. 509–511) ‘And I am tottering with hunger Such that I thought we had been lying down Very close to England’
As for the time represented on stage, the use of Latin liturgical language for the purposes of expressive swearing establishes the zeitgeist(17): (17)
1 Pastor. Ressurrex a mortruus! Haue hold my hand. Iudas carnas dominus! I may not well stand. My foot sleeps, by Jesus! (“Secunda Pastorum,” ll. 504–508) ‘First Shepherd. He rose from the dead!. Have hold of my hand. Judas flesh Lord! I may not well stand. My foot sleeps, by Jesus!’
573
574
IV Middle English As Stevens and Cawley (1994, Volume 2) note, garbled or corrupted Latin informs lines 504 and 506. Would the irony of ‘Judas flesh Lord!’ or ‘Judas, lord of the flesh!’ as a corruption of the hymn, laudes canas domino, ‘sing praises to the Lord’, be a source of humor for laity as well as clergy watching the performance? Swearing by Jesus in a Nativity play in which Jesus has not yet been born should have garnered a laugh, even while the appropriateness of the utterance for people used to swearing – outside of churches, of course – would let the audience identify with the 1st-century shepherds. This pattern of merging the past action of the narrative with the present of the audience repeats itself later on. Arriving at the stable where the baby Jesus lies, the Third Shepherd addresses him (18): (18)
Hayll, swete is thy chere! My hart wold blede To se the sytt here In so poore wede, With no pennys (“Secunda Pastorum”, ll. 1054–1058) ‘Hail, sweet is your cheer! My heart bleeds, To see you sit here In so poor clothes, With no pennies’.
If pennies were the coin of the realm in England, so, too, they must have been in Nazareth, thirteen hundred years earlier. Moments before offering his gift, the Second Shepherd tells the baby that he wold drynk on thy cop ‘would drink from your cup’, a 14th-century allusion to the chalice from which Christians will drink, when, in the first century setting of the play, the Eucharist has not been established (“Secunda Pastorum”, l. 1048).
4 Summary Colloquial expression, anachronism, and prosody give the Wakefield pageants a style as realistic and profane as it is sacred. Its easy use of spoken vernacular stands in opposition to the unidiomatic language representing Latin in certain translations. In between these extremes is the living mixture of usages and styles we know as ME writing. If we accept the notion that a self-consciously literary language does not manifest itself in English until very late in the medieval period, we nevertheless find its groundwork lain in the style of some of the writing reviewed here. While the literary language of Old English may not have died entirely with the Conquest, being maintained in a different guise in the alliterative poetry and prose of Middle English, we acknowledge the extent of the breach in the many translations from Latin and French that comprise English writing. Let me conclude by returning to the problematic question of ME literary language. We can classify texts by time period and genre and recount stylistic features. But we must take care in what we abstract from style, for any conclusions we reach must be evaluated with the medieval context. Finally we may do well to ask, if the medieval world itself does not understand “literature” in the same way we do, how productive is it for us to force our modern, disciplinary categories on its writings?
35 Middle English: Literary language
5 References Andrew, Malcolm and Ronald Waldron (eds.). 2002. The Poems of the Pearl Manuscript. 4th edn. Exeter: University of Exeter Press. Arnovick, Leslie K. 2006. Written Reliquaries: The Resonance of Orality in Medieval English Texts. Amsterdam/Philadephia: John Benjamins. Babington, Churchill. 1869. Polychronicon Ranulphi Higden monachi Cestrensis; together with the English Translations of John Trevisa and of an Unknown Writer of the Fifteenth Century. London: Longman. Benson, Larry D. 1965. Art and Tradition in Sir Gawain and the Green Knight. New Brunswick: Rutgers University Press. Benson, Larry D. (ed.). 1987. The Riverside Chaucer. Boston: Houghton Mifflin. Blake, Norman F. 1969. Rhythmical alliteration. Modern Philology 67: 118–124. Blake, Norman F. 1977. The English Language in Medieval Literature. London: Dent. Blake, Norman F. 1987. Late medieval prose. In: W. F. Bolton (ed.), The Middle Ages, 369–399. New York: Peter Bedrick Books. Blake, Norman F. 1992. The literary language. In: Norman F. Blake (ed.), The Cambridge History of the English Language, Vol. II. 1066–1476, 500–541. Cambridge: Cambridge University Press. Brook, G. L. and R. F. Leslie (eds.). 1963. Laʒamon: Brut. (Early English Text Society, 277.) London/New York: Oxford University Press. Burnley, David. 1983. A Guide to Chaucer’s Language. Houndmills: Macmillan. Burnley, David and Matsuji Tajima (eds.). 1994. The Language of Middle English Literature. Cambridge: D. S. Brewer. Burrow, J. A. and Thorlac Turville-Petre. 2005. A Book of Middle English. 3rd edn. Malden: Blackwell. Cannon, Christopher. 1998. The Making of Chaucer’s English: A Study of Words. Cambridge: Cambridge University Press. Cannon, Christopher. 2008. Middle English Literature: A Cultural History. Cambridge: Polity Press. Chism, Christine. 1999. Romance. In: Scanlon (ed.), 57–69. Cooper, Helen. 2010. Prose romances. In: Edwards (ed.), 215–229. David, Alfred and James Simpson (eds.). 2006. The Norton Anthology of English Literature. Vol. A. New York and London: W.W. Norton & Company. Davis, Norman. 1987. Language and Versification. In Benson (ed.), xxix–xiv. Edwards, A. S. G. (ed.). 2010. A Companion to Middle English Prose. Cambridge: D.S. Brewer. Ginsberg, Warren S. 1987. Notes to The Clerk’s Prologue and Tale. In Benson (ed.), 879– 884. Hammond, Gerald and Sylvia Adamson (eds.). 1996–2011. Bible in English. Chadwyck-Healey. http://collections.chadwyck.co.uk/marketing/products/about_ilc.jsp?collection=bie Hanna, Ralph. 1999. Alliterative poetry. In: Scanlon (ed.), 488–512. Hanna, Ralph. 2010 [2004]. Rolle and related works. In: Edwards (ed.), 19–31. Hartung, Albert E. and J. Burke Severs (eds.). 1969–93. A Manual of the Writings in Middle English. 9 vols. Hamden, CT: Anchor Books. Hasenfratz, Robert (ed.). 2000. Ancrene Wisse. Kalamazoo: Medieval Institute Publications. Horobin, Simon. 2003. The Language of the Chaucer Tradition. Cambridge: D. S. Brewer. Horobin, Simon. 2007. Chaucer’s Language. Houndmills: Palgrave Macmillan. Kelly, Henry Ansgar. 2010. The Englishness of the Wycliffite Bible. Paper delivered to the Medieval Association of the Pacific, Tacoma, March 2010. Kolve, V. A. 1966. The Play Called Corpus Christi. Palo Alto: Stanford University Press. Macaulay, G. C. 1900–01. The English Works of John Gower. (Early English Text Society, E.S. 81–82.) London: K. Paul, Trench, Tru¨bner and Co.
575
576
IV Middle English Minkova, Donka. 2003. Alliteration and Sound Change in Early English. Cambridge: Cambridge University Press. Minkova, Donka. 2007. The forms of verse. In: Peter Brown (ed.), A Companion to Medieval English Literature and Culture c. 1350–1500, 176–193. Chichester: Wiley-Blackwell. Newhauser, Richard. 2009. Religious writing. In: Scanlon (ed.), 37–55. Pearsall, Derek. 1977. Old English and Middle English Poetry. London: Routledge and Kegan Paul. Renoir, Alain. 1988. A Key to Old Poems. University Park: The Pennsylvania State University Press. Scanlon, Larry (ed.). 1999. The Cambridge Companion to Medieval English Literature 1100–1500. Cambridge: Cambridge University Press. Simmons, Thomas Frederick (ed.). 1879. The Lay Folks Mass Book. London: Tru¨bner and Co. Smith, Jeremy J. 2009. Writing in English, writing in England: Language. In: Marilyn Corrie (ed.), A Concise Companion to Middle English Literature, 145–165. Chichester: Blackwell Publishing Ltd. Spearing, A. C. 2010. Margery Kempe. In: Edwards (ed.), 83–97. Stevens, Martin and A. C. Cawley (eds.). 1994. The Towneley Plays. 2 vols. Oxford: Oxford University Press. Tolkien, J. R. R. 1929. Ancrene Wisse and Hali Meiðhad. Essays and Studies 14: 104–126. Tolkien, J. R. R. and E. V. Gordon (eds.). 1967. Sir Gawain and the Green Knight. Revised by Norman Davis. Oxford: Oxford University Press. Turville-Petre, Thorlac. 1996. England the Nation: Language, Literature, and National Identity, 1290–1340. Oxford: Clarendon Press. Watson, Nicholas. 1999. The politics of Middle English writing. In: Jocelyn Wogan-Browne, Nicholas Watson, Andrew Taylor, and Ruth Evans (eds.), The Idea of the Vernacular, 331–352. University Park: Pennsylvania State University Press.
Leslie K. Arnovick, Vancouver (Canada)
36 Middle English: The language of Chaucer 1 2 3 4
Spelling and dialect Grammar Vocabulary References
Abstract This chapter provides an introduction to the variety of Middle English employed by the 14th-century poet Geoffrey Chaucer, situating it within the development of the English language in an important stage in its development. The chapter begins by outlining the process of standardization, considering the place of Chaucer’s language in the emergence of a standard variety of written English from the numerous competing written varieties of Middle English. Section 2 provides an overview of Chaucer’s noun, verb, and pronoun systems, highlighting important differences between Chaucer’s usage and those that preceded and followed him. Section 3 focuses on Chaucer’s vocabulary, its origins, and the impact of external sources, principally French and Latin, as well as the various Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 576–587
36 Middle English: The language of Chaucer methodologies and electronic resources available for assessing Chaucer’s exploitation of the richness of the Middle English lexicon.
1 Spelling and dialect The ME period was a time of considerable linguistic variation. There was no standard written or spoken variety of the language and, as a consequence, writers simply wrote using their native dialect. Chaucer was a Londoner by birth and so the dialect he used was that of London. This should not be taken to imply that the London dialect was of a higher status than other regional varieties; the author of the alliterative poem Sir Gawain and the Green Knight was of Cheshire origin and so wrote his works using that dialect. Speakers of Middle English were certainly aware of regional differences in pronunciation, as may be seen in Chaucer’s characterization of the Northern dialect in the Reeve’s Tale. Here Chaucer employs certain characteristic features of Northern pronunciation, grammar, and vocabulary as part of his depiction of two Cambridge undergraduates who are of Northern origins. It is tempting for modern readers to see this as early evidence of north-south prejudices that are found among presentday English speakers, but the lack of evidence for a standard variety makes such an interpretation unlikely. Furthermore, the fact that the Northerners are Cambridge undergraduates suggests a higher status for the Northern dialect than the Southern speech used by the cheating Miller. Since J. R. R. Tolkien’s (1934) important study of Chaucer’s use of Northern dialect in the Reeve’s Tale there have been numerous studies of Chaucer’s depiction of dialect, although scholars continue to struggle to explain why Chaucer should have chosen to employ dialect speech in this particular tale and with these particular characters. For an overview of discussion of this issue and a new interpretation see Machan (2003). While there is no evidence of a standard spoken variety of Middle English, there is, however, some evidence indicating the emergence of a standard written language towards the end of the ME period (see Schaefer, Chapter 33). This development began during the late 14th century, but it is important to recognize that this process was not influenced by Chaucer, nor was he directly responsible for ensuring the establishment and adoption of this standard variety. Throughout much of the ME period spelling was considerably more varied than in Present-day English, as a result of its lack of status and its local functions. However, by the end of the ME period, the vernacular began to be used for communication on a wider scale and this kind of variation became inefficient. As a result there was an increased need for a standardized written variety of Middle English which could be understood over a wide geographical area. In a seminal discussion of the development of London English during the 14th and 15th centuries, M. L. Samuels (1963) distinguished four “types” of written standard which he labelled types I–IV. Type I, also known as the Central Midlands Standard, is found in a number of texts associated with John Wycliffe and the Lollard movement. This language is found in a large number of manuscripts of religious texts and Bible translations produced by the Lollards, copied and circulated widely throughout the country. This type of language is based upon the dialects of the Central Midlands counties, and characteristic forms include sich ‘such’, mych, ‘much’, ony ‘any’, silf ‘self’, stide ‘stead’, ʒouun ‘given’, siʒ ‘saw’. Samuels’ Type II is found in a group of manuscripts copied in London in the mid to late 14th century; a group which includes the Auchinleck
577
578
IV Middle English manuscript. The Auchinleck manuscript was produced in London around 1340 by six scribes, some of whom were Londoners while others were natives of the West Midlands. Characteristic Type II features include forms which are common to the Norfolk and Suffolk dialects, and are thought to derive from immigration into the capital from those counties. Examples of such forms are þai, hij ‘they’, þeiʒ ‘though’, werld ‘world’, þat ilch(e), ilch(e) ‘that very’. The London dialect changed towards the end of the 14th century, as a result of continued immigration. This late 14th- and early 15th-century variety, termed by Samuels Type III, is the dialect of the earliest Chaucer manuscripts: the Hengwrt and Ellesmere manuscripts of the Canterbury Tales and Corpus Christi College, Cambridge MS 61 of Troilus and Criseyde. Other Type III documents are early civic documents such as the London guild returns and the Petition of the Mercers’ Company, as well as the early 15th-century holograph manuscripts containing the works of the poet Thomas Hoccleve. Characteristic features of this language are forms common to the Central Midlands dialects, including they ‘they’, hir(e) ‘their’, though ‘though’, yaf ‘gave’, nat ‘not’, swich ‘such’. Type III was subsequently replaced around 1430 by Type IV, also termed by Samuels “Chancery Standard” which was employed by government clerks working in the Chancery, Signet, and Privy Seal Offices. Type IV also shows the influence of forms found in the Central Midlands dialects although these have been further supplemented by forms originally restricted to the North Midlands, e.g. theyre ‘their’, thorough ‘through’, such(e) ‘such’, gaf ‘gave’, not ‘not’. Michael Benskin’s extensive analysis of Chancery and other administrative documents copied throughout the 15th century has led him to question the status of Chancery Standard and the claims made by John Fisher concerning its early adoption among Chancery scribes (Benskin 2004; Fisher 1996) What is clear, however, is that while Chaucer’s dialect may have played an important role in the standardization of written English, it was not ultimately the ancestor of the standard language. It is important to be aware that these types are not fixed standards, with fixed sets of rules from which deviation was stigmatized, as is the case with our PDE standard written language. It is more accurate to see these four types as focused standards, in that texts written using one of these types share a number of distinctive spelling features, but not all of them will appear in all texts (Smith 1996). So while PDE standard written English does not permit variation, these standardized, or focused, types do tolerate a certain amount of variation. So Chaucer’s dialect belongs to a variety known to scholars as Type III London English, a standardized usage employed in a number of texts copied in London towards the end of the 14th and beginning of the 15th centuries. These labels have, however, come under considerable scrutiny since Samuels’ article was first published in 1963. Horobin (2003) has studied the variation found within documents copied in Type III and Type IV, arguing that the distinction between these types is more a reflection of differences of register rather than date. The majority of manuscripts written in Samuels’ Type III contain literary texts, particularly ones connected with Chaucer, suggesting that it may have functioned as a kind of literary, or Chaucerian, standard language. More recently, palaeographical analysis has further complicated the picture, by showing that many Type III documents were in fact copied by a single scribe, thereby undermining their claim to be representative of a standard language (Horobin and Mooney 2004).
36 Middle English: The language of Chaucer
579
2 Grammar The process of simplification and loss of inflectional endings that is characteristic of Middle English (see Wełna, Chapter 37) is comparatively well advanced by the time of Chaucer’s writing. This means that many of the special endings that were employed to indicate case, number, and gender in Old English were no longer in use. This is apparent from a comparison of Chaucer’s noun paradigm with the equivalent OE version (see Table 36.1). Table 36.1: Declension of stone in Chaucer and in Old English Old English NOM ACC GEN DAT
Chaucer’s language
SG
PL
SG
PL
sta¯n sta¯n sta¯nes sta¯ne
sta¯nas sta¯nas sta¯na sta¯num
stoon stoon stoones stoon(e)
stoones stoones stoones stoones
Where Old English has distinct endings to mark the nominative/accusative, dative, and genitive singular and plural, Chaucer’s system relies almost entirely on an endingless form and an ending which indicates plurality and possession. There is one further ending, , used to mark the dative singular, although this is only found in fossilized phrases such as on honde. Adjective inflection shows a similar tendency towards syncretism and loss, with the result that the only ending used in Chaucer’s language is -e. This ending is added to plural adjectives, as in stronge men, and to singular adjectives that appear after a determiner, as in the olde man, this yonge sonne. The only adjectives that do not take this inflection, therefore, are ones that are not preceded by a determiner, as in the man is old, or an old man (Burnley 1982). This distinction had already been lost in the Northern dialects of Middle English, but it is preserved by a number of Southern scribes. However, the consistency with which these distinctions are maintained throughout Chaucer’s works is unusual and presumably reflects a conscious desire to preserve this increasingly archaic feature for metrical reasons. As in Present-day English, ME pronouns were inflected according to number, case, and gender. Third person pronouns were selected according to natural gender, i.e. real world sex distinctions, as in Present-day English, rather than according to grammatical gender as found in Old English. The forms of the personal pronouns used by Chaucer are as shown in Tables 36.2 and 36.3. Table 36.2: First and second person pronouns in Chaucer 1P
SG
PL
NOM
I (ich, ik) me my(n)(e) me
we us our(e)(s) us
ACC GEN DAT
(Continued )
580
IV Middle English Table 36.2: Continued 1P
SG
PL
NOM
thou/thow thyn(e) thee
ye your(e)(s) you/yow
GEN DAT
Table 36.3: Third person pronouns in Chaucer 3P SG
MASC
FEM
NEUT
3P PL
all genders
NOM
he hym, him his hym, him
she hir(e), hyr(e) hir(e)(s) hir(e), hyr(e)
it, hit it, hit his it, hit
NOM
they hem hir(e)(s) hem
ACC GEN DAT
ACC GEN DAT
The above tables show that there was no formal distinction between the accusative and dative cases in Middle English; these two cases had merged so that the same pronoun was used for both. The different forms of the first person singular nominative pronoun, I, ich, and ik, represent dialect distinctions: the southern form ich is found as an alternative throughout Chaucer’s works, while Ik is a variant employed only in the Reeve’s Prologue as part of the Norfolk dialect attributed to that character. Another difference between Middle English and Present-day English in the above table is the two forms of the second person pronoun: thou/ye, equivalent to PDE you. This number distinction is inherited from Old English, but in Middle English it took on additional functions through contact with French (cf. Busse, Chapter 46). French maintains a pragmatic distinction in the use of the singular and plural pronouns when used to address a single individual. The plural vous form is used to indicate respect and formality, while the singular form, tu, is reserved to signal familiarity or a lack of respect. A similar distinction is found in Chaucer’s use of thou and ye, and an understanding of this distinction is important in appreciating the subtleties in the shifting relations between characters in Chaucer’s works. It was conventional for courtly men and women to address each other using the plural pronoun, as is apparent in the following example from the Franklin’s Tale in which Dorigen consents to be Arveragus’ wife, as in (1): (1)
“Sire, sith of youre gentillesse Ye profre me to have so large a reyne, Ne wolde nevere God bitwixe us tweyne, As in my gilt, were outher werre or stryf. Sire, I wol be youre humble trewe wyf; Have heer my trouthe – til that myn herte breste.” (F 754–759; Chaucer quotations following Benson [ed.] 1988) ‘ “Sir, since by your nobility you allow me to have so loose a rein, God would never wish there to be either war or strife between us two through my fault. Sir, I am willing to be your humble, true wife. Receive here my pledge – until me heart breaks” ’
36 Middle English: The language of Chaucer
581
The use of the second person plural pronoun here is intended to indicate Dorigen’s respect for her husband, and is all the more striking given that the terms of the marriage arrangement are that both husband and wife are to be equals. While the plural pronoun is used to indicate respect between individuals, the singular form was employed to indicate disrespect or contempt, as in the Host’s abrupt interruption of Chaucer’s Tale of Sir Thopas, as in (2): (2)
“Namoore of this, for Goddes dignitee,” Quod oure Hooste, “for thou makest me So wery of thy verray lewednesse” (B2 2109–2111) ‘ “No more of this, for God’s sake”, said our Host, “for you make me so fed up by your complete ignorance” ’
However, the singular form may also be used to express familiarity and intimacy. For instance, in Chaucer’s courtly romance Troilus and Criseyde the two lovers address each other using the plural pronoun throughout the poem, although both switch to the singular pronoun at moments of high emotional intensity, such as Troilus’ pledge of commitment to Criseyde: “For I am thyn, by God and by my trouthe!” (3.1512). While these examples show Chaucer’s subtle exploitation of the pragmatic distinctions between singular and plural pronouns, there are occurrences where Chaucer pays less attention to such nuances of usage. The neuter form of the third person singular pronoun remained identical with the masculine form his, as in Old English. The replacement of this form with its PDE equivalent its did not occur until the EModE period. However, the third person plural pronouns are quite different from those used in Present-day English. This is because Chaucer’s system represents a transitional stage in the replacement of the OE pronouns hie, hiera, him with forms derived from Old Norse (ON): they, their, them. These ON third person plural pronouns were adopted during the ME period; presumably in an attempt to avoid confusion with other personal pronouns. As with many ON loanwords, these pronouns are first recorded in the Northern and East Midlands dialects and only later appear in the Southern dialects. It is not until the 14th century that we see these forms appearing in texts copied in London, and Chaucer used only the nominative form they, alongside the forms hem and hir derived from Old English. Chaucer’s verb endings show a number of differences from those of Present-day English, as may be seen in the Table 36.4. Table 36.4: Present tense verbal endings in Chaucer and in Present-day English
1P SG 2P SG 3P SG PL
Chaucer’s language
Present-day English
loue louest loueth loue(n)
love love loves love
While the ending is the main form for the 3rd person singular present indicative in Chaucer’s writings, another ending in is also found in a handful of examples.
582
IV Middle English As with the ON derived third person plural pronouns this ending was originally a Northern feature, only adopted into London English during the 14th century. Chaucer evidently knew the ending, as he used it as a variant in rhyme twice in the Book of the Duchess, (3), but it was not a consistent feature of his own dialect: (3)
That never was founden, as it telles, Bord ne man, ne nothing elles. (lines 73–74) ‘That never was found, as it is told, plank nor man, nor nothing else’.
The only other instance of its use in Chaucer’s works is in the dialect speech in the Reeve’s Tale, indicating clearly that Chaucer considered this verb ending to be a northern dialect feature.
3 Vocabulary A large proportion of Chaucer’s vocabulary is made up of words of OE origin, and these tend to be words which constitute the core of his vocabulary: high frequency items which refer to everyday concepts. Chaucer’s vocabulary also comprises a substantial number of words of ON derivation, the majority of which represent core concepts such as housbonde, law, sky, skile, though. We saw in the previous section that ON words were adopted earliest in the more northerly dialects of Middle English, and it is therefore not surprising that Chaucer prefers words of OE derivation such as swich ‘such’, chirch ‘church’, ey ‘egg’ to the ON equivalents slik, kirk, egg. We have already seen that Chaucer employed only the nominative third person plural ON pronoun form, they, alongside hem, not them and hir/her, not their. The only appearance of the forms them and their in Chaucer’s works is in the Northern dialect of the students in the Reeve’s Tale, indicating that Chaucer considered these forms to be northern. Chaucer’s lack of use of these pronouns is striking, given that they replaced the OE forms in the London dialect shortly after his death. While he generally preferred OE words rather than ones of ON descent, Chaucer did employ some ON words alongside their OE equivalents, as in the cases of thilke/same; give/yive; against/ayeinst. There are some examples of doublets of this kind where there was evidently a distinction in meaning or register which can be important for our understanding of Chaucer’s work. A good example of this concerns Chaucer’s use of the words cherl, which is of OE origin, and carl, a borrowing from Old Norse. Both are used by Chaucer with reference to a peasant, although it is apparent from the contexts within which they are used, and the people to which the terms are applied, that carl was considered to be a considerably more contemptuous term of abuse than cherl (Burnley 1989: 50–51). Discussions of Chaucer’s vocabulary have tended to focus most on his borrowings from French, and principally on those French words that are first recorded in his works. This approach is most evident in Joseph Mersand’s book Chaucer’s Romance Vocabulary, which calculates that Chaucer’s complete vocabulary was 8,072 words, of which 4,189 are derived from Romance sources. Mersand drew on the evidence of first citations in the Oxford English Dictionary (OED) (Simpson and Weiner [eds.] 1989) to calculate that 1,180 of the Romance words used by Chaucer first appear in English in Chaucer’s works, and he argued that these words represented Chaucer’s “gifts to the English language” (Mersand 1937: 56). But the methodology that Mersand
36 Middle English: The language of Chaucer employed, and the figures that he produced, are based upon a number of problematic assumptions. The fact that the earliest recorded use of a word in the OED is found in Chaucer’s works does not necessarily mean that Chaucer was the first writer to use it. As a major literary writer Chaucer’s works were comprehensively mined for quotations by the editors of the OED, and so it is hardly surprising that his works are well represented in its quotations. By contrast many of his contemporaries’, and predecessors’ works were less available in modern editions and consequently less widely read for the dictionary, with the result that Chaucer’s uses often appear as the earliest recorded occurrences. The recent completion of the Middle English Dictionary (MED) (Kurath et al. 1952–2001), with a far more comprehensive treatment of the ME period, reveals that many of the words cited as first occurring in Chaucer’s works in OED are in fact found in earlier texts. A further theoretical objection to Mersand’s approach is the way that he assumed that the earliest recorded use of a word is likely to represent its first appearance in English. Many loanwords would have been used in speech for some time before they appeared in the written language, while it is also likely that earlier written instances have been lost. So the fact that the first attestation of a word appears in Chaucer’s works does not necessarily indicate that he was responsible for introducing it into English. Another objection to Mersand’s approach is the assumption that the borrowing of words from French and Latin was a novelty, ignoring the fact that earlier ME writers made considerable use of this same practice. Christopher Cannon (1998) has reconsidered Mersand’s approach, highlighting many of its shortcomings. Yet Cannon’s book suffers from similar limitations in the ways that it draws upon lexicographical and etymological information. Having discussed the flaws in Mersand’s approach, Cannon attempts to show how dictionary evidence can be employed to determine the status of Chaucer’s words. In a discussion of Chaucer’s colloquial style Cannon considers how dictionaries can supply evidence of words derived from the spoken language. He argues that by identifying gaps in the lexicographical record we can distinguish words which Chaucer introduced from the spoken into the written language. According to Cannon, where a word appears in Old English and then does not appear in Middle English before Chaucer, it is likely to represent a colloquialism which fell out of use in the written language and survived only in speech. Using this methodology Cannon produces a list of 36 words which he describes as “resolutely native colloquialisms” (Cannon 1998: 161). Yet Cannon allows the dictionary evidence to stand on its own and he does not proceed to analyze the words themselves, and the use Chaucer made of them. In fact the words listed by Cannon seem most unlikely to be colloquialisms, and the simplest explanation for their absence from the written record is their technical or specialized meanings. For example Cannon’s list includes the word harpe-stringe, a word which perhaps unsurprisingly is used just once by Chaucer. But the lack of occurrences of this word prior to its appearance in the House of Fame tells us very little, apart from the fact that harp-strings are not much discussed in early Middle English writing. The frequent appearance of both harp and string in the early Middle English record further undermines any attempt to assign the word harpe-stringe to a colloquial register. Other words on the list are equally technical and unusual, such as ordal ‘trial by ordeal’, thurrok ‘the bilge of a ship’, last ‘a ship’s load’, and chimb ‘the rim of a barrel’, suggesting that their lack of earlier attestation is simply due to their specialized status.
583
584
IV Middle English It is clear from the discussion so far that trying to base our assessment of Chaucer’s lexicon on dictionary evidence is limited in a variety of ways. Simply counting up French borrowings, or ones first attested in Chaucer’s works, tells us little about the status of such words in Chaucer’s language. To develop a fuller appreciation of the function of borrowing in Chaucer’s works we must make a distinction between French words used by Chaucer which were already in use in English, and those which were more recent borrowings. This distinction is apparent from Chaucer’s discussion of the word fruit in the Parson’s Tale, where he refers to it as the English equivalent of the Latin fructus (I 869). The word fruyt is ultimately of French origin, though it is recorded in Middle English from the early 13th century, and by Chaucer’s time it was clearly well assimilated into the English lexicon. A similar example is found in the Second Nun’s Tale, where peple is given as the English word for Greek leos (G 106). Peple is also a French loanword, but it is one that is found in Middle English from the late 13th century and thus well established within the English lexicon by the late 14th century. One way of trying to distinguish between recent loans and those that had been fully assimilated into English by the time Chaucer used them is to consider their distribution among texts written prior to and contemporary with Chaucer. Where a French loanword is first attested early in the ME period, and found in a variety of different sources by the time Chaucer used it, we can be confident that Chaucer’s use of it would not have been considered novel. In employing such an approach we must remain conscious of the caveats regarding first attestations noted above, as well as the patchy nature of the surviving written record. When we begin to analyze Chaucer’s use of Romance loans it becomes immediately apparent that there is a distinction between Romance words that fill a lexical gap, and words introduced alongside existing English words as stylistic alternatives. The dominance of the French and Latin languages throughout much of the early ME period meant that there were no English words for many key concepts, creating the need for groups of new lexical items. Chaucer’s Treatise on the Astrolabe, for example, is one of the earliest ME scientific works. As astronomical texts were traditionally written in Latin in the Middle Ages, it is not surprising that Chaucer was compelled to introduce a number of Latin loanwords into English to provide him with the technical terms he needed. So Chaucer’s Astrolabe contains what appear to be the first attestations of technical words such as almicanteras, azimutz, consentrik, and declinacioun. That Chaucer expected his audience to consider such words to appear foreign is apparent from his tendency to gloss them, as shown by the way he introduces the word azimutz: “these same strykes or diuisiouns ben clepid azimutz” (1.19). A similar example is found in the introduction of the technical term altitude, which Chaucer carefully glosses when it is first introduced, but then expects his readers to be able to understand it on subsequent occasions: “Thyn Astrolabie hath a ring to putten on the thombe of thi right hond in taking the height of thinges. And tak kep, for from henes forthward I wol clepen the heighte of any thing that is taken by the rewle ‘the altitude’, withoute moo wordes” (1.1). Where Chaucer differs from other writers in their use of technical terms is in employing such words in a wider range of contexts. For example the astronomical term declinacioun appears in the Franklin’s Tale in the following context, (4):
36 Middle English: The language of Chaucer (4)
He seyde, “Appollo, god and governour Of every plaunte, herbe, tree and flour, That yevest after thy declinacioun, To ech of hem his tyme and his seson, (F 1031–1034) ‘He said, “Appollo, god and governor of every plant, herb, tree and flower, who gives to each of them their time and season, according to your distance from the equinoctial’
Alongside the borrowing of technical terms to provide vernacular equivalents to words available only in French or Latin, Chaucer’s works also employ a number of French words where English equivalents were already available. The introduction of such words was driven by stylistic factors, reflecting the high status attached to French culture during the late Middle Ages, especially within the English court. French loanwords comprise a large proportion of Chaucer’s words concerned with courtly concepts such as gentillesse, curteisye, chivalrye, honour, pitee, mercy. Another way of assessing a word’s status in Chaucer’s works is to examine all of its instances across his oeuvre. An excellent model for this kind of approach is a study by E. T. Donaldson (1970) which demonstrated Chaucer’s ironic use of the courtly love language of earlier romances and lyrics. Donaldson considered the distribution of a group of words that are generally restricted to fabliaux contexts, such as the word hende, meaning ‘noble, courtly, or refined’. This word is found almost exclusively in the Miller’s Tale, where it is employed as an epithet for the Oxford student “hende Nicholas”. There are two occurrences of the word outside the Miller’s Tale: one in the Wife of Bath’s Prologue, where it is used of another Oxford student of dubious morals, the Wife of Bath’s fifth husband Jankyn. The second instance is its use by Harry Bailly in his efforts to settle a dispute between the Friar and the Summoner (D 1286–1287). This distribution of the word hende indicates that, while its meaning may have been ‘noble, courtly, or refined’, Chaucer only ever used the word ironically. Studies of this kind need to be extended to cover all fields of Chaucerian discourse and stylistic registers, while making use of the considerable resources that are now available for studying the distribution of individual words across Chaucer’s works and the whole ME corpus. The importance of reconstructing the “architecture” of Chaucer’s language to identify the various contextual factors that relate to individual word choice has been demonstrated by David Burnley (1979, 1989). An effective demonstration of this may be seen in Burnley’s discussion of groups of synonyms such as hyre, guerdoun, mede, all with the core meaning ‘reward’ (Burnley 1989: 203–213). Burnley’s discussion of words used by Chaucer to refer to the ‘heart’ identified a technical usage in the word corage, generally used to translate the Latin word animus and referring to the rational part of the soul. This sheds interesting light on Chaucer’s use of the word in the famous depiction of Criseyde as “slydynge of corage” (Troilus 5.825) (Burnley 1989: 216–217). The resources available for this kind of study have been greatly enhanced in recent years so that it is now possible to carry out detailed analyses of the occurrences of individual words across large quantities of text. A valuable resource is the Glossarial DataBase of Middle English, based on the texts edited in the Riverside Chaucer, which allows the identification of each occurrence of a particular lexeme across the whole of the Canterbury Tales. Searches can also be limited to particular parts of speech,
585
586
IV Middle English enabling more fine-grained analysis. Searching the Glossarial DataBase (Benson 2006) enables the identification of patterns in the distribution of certain words across Chaucer’s works of the kind observed by Donaldson, discussed above. It also allows us to observe whether there are any contextual factors that might govern a particular word’s use, such as whether it is spoken by a particular character, in a particular text-type, in a prose text, in rhyming position etc. As well as analyzing the distribution of individual words across Chaucer’s works, we can also examine their occurrence in other Middle English works by searching electronic corpora. The Middle English Compendium (McSparran [ed.] 2006) includes a corpus of Middle English verse and prose which contains texts of some 60 Middle English works which can be searched in their entirety, or by single texts or groups of texts, and which includes complete texts of Chaucer’s Canterbury Tales, Troilus and Criseyde, Boece, and Treatise on the Astrolabe. While much important work has been done on Chaucer’s language (see also Horobin 2011), there is considerably more that needs to be done. Past studies have tended to focus too much on lexicographical evidence rather than analyzing the words themselves and the contexts within which they are used. There has also been a tendency to consider Chaucer’s usage in a vacuum, rather than considering his language in relation to that of his contemporaries. A renaissance of interest in the study of Chaucer manuscripts has also led to a reassessment of Chaucer’s language in the light of the surviving textual evidence, forcing scholars to reconsider the classifications that have been accepted since Samuels’ influential paper of 1963. The transcription of these texts into electronic form by the Canterbury Tales project (Robinson 1996; Solopova 2000) and the existence of other electronic editions is further facilitating this reconsideration of the textual evidence, enabling large scale analyses of details of spelling and morphology that were impossible in the pre-digital age.
4 References Benskin, Michael. 2004. Chancery Standard. In: Christian J. Kay, Carole Hough, and Irene´ Wotherspoon (eds.), New Perspectives on English Historical Linguistics: Selected Papers from 12 ICEHL. Vol. II: Lexis and Transmission, 1–40. Amsterdam/Philadelphia: John Benjamins. Benson, Larry D. (gen. ed.) 1988. The Riverside Chaucer. Oxford: Oxford University Press. Benson, Larry D. 2006. A Glossarial DataBase of Middle English. http://www.hti.umich.edu/g/ gloss/ Burnley, J. D. 1979. Chaucer’s Language and the Philosophers’ Tradition. Cambridge: D. S. Brewer. Burnley, J. D. 1982. Inflexion in Chaucer’s adjectives. Neuphilologische Mitteilungen 83: 169–177. Burnley, J. D. 1989. The Language of Chaucer. Basingstoke: Macmillan. (Reprint of A Guide to Chaucer’s Language. Basingstoke: Macmillan, 1983.) Cannon, Christopher. 1998. The Making of Chaucer’s English: A Study of Words. Cambridge: Cambridge University Press. Donaldson, E. Talbot. 1970. Idiom of popular poetry in the Miller’s Tale. In: Speaking of Chaucer, 13–29. London: Athlone. Fisher, John H. 1996. The Emergence of Standard English. Lexington, KY: The University Press of Kentucky. Horobin, Simon. 2003. The Language of the Chaucer Tradition. Cambridge: D. S. Brewer. Horobin, Simon. 2011. Chancer and late medieval language. Literature Compass 8(5): 258–265.
36 Middle English: The language of Chaucer
587
Horobin, Simon and Linne Mooney. 2004. A Piers Plowman manuscript by the Hengwrt/Ellesmere scribe and its implications for London Standard English. Studies in the Age of Chaucer 26: 65–112. Kurath, Hans, Sherman M. Kuhn, John Reidy, and Robert E. Lewis. 1952–2001. Middle English Dictionary. Ann Arbor: University of Michigan Press. http://quod.lib.umich.edu/m/med/ Machan, Tim William. 2003. English in the Middle Ages. Oxford: Oxford University Press. McSparran, Frances (ed.). 2006. Middle English Compendium. Ann Arbor: University of Michigan Press. http://quod.lib.umich.edu/m/mec/ Mersand, Joseph. 1937. Chaucer’s Romance Vocabulary. New York: Comet. Simpson, John A. and Edmund S. C. Weiner (eds.). 1989. Oxford English Dictionary. 2nd edn. Oxford: Oxford University Press. Robinson, P. M. W. (ed.). 1996. The Wife of Bath’s Prologue on CD-ROM. Cambridge: Cambridge University Press. Samuels, M. L. 1963. Some applications of Middle English dialectology. English Studies 44: 81–94. (Reprinted in Margaret Laing [ed.], Middle English Dialectology: Essays on Some Principles and Problems, 64–80. Aberdeen: Aberdeen University Press.) Smith, Jeremy J. 1996. An Historical Study of English: Function, Form and Change. London: Routledge. Solopova, Elizabeth (ed.). 2000. The General Prologue on CD-ROM. Cambridge: Cambridge University Press. Tolkien, J. R. R. 1934. Chaucer as a philologist: The Reeve’s Tale. Transactions of the Philological Society (1934), 1–70.
Simon Horobin, Oxford (UK)
V Early Modern English 37 Early Modern English: Phonology 1 2 3 4 5 6 7
Introduction Word stress Syllable reduction Consonants Vowels Summary References
Abstract In the Early Modern English (EModE) period, English underwent a number of substantial changes in all phonological subsystems, which transformed the Middle English system into a distinctly modern one. The present chapter highlights in turn changes in lexical stress patterns, the reduction of unstressed syllables, changes in the distribution of certain consonants and their allophones (in particular /h/, [c¸], [x], /r/), the reduction of consonant clusters (including the emergence of the novel /ʒ / and /ŋ / phonemes) and changes in the vowel system. A large part of the chapter is devoted to the important shifts undergone by the latter category of sounds in the Early Modern era; yet it excludes the massive turnover known as the Great Vowel Shift (which is treated in Krug, Chapter 48). The vowel changes are, for expository purposes, subdivided into unconditioned and conditioned changes; the subsystems of long vowels, diphthongs, and short vowels are treated separately. The chapter describes the most important qualitative changes in long vowels (beyond the Great Vowel Shift), the monophthongization of many ME diphthongs, the development of some of the short vowels, and the interfering effect of certain consonantal environments, partly leading to phonemic splits.
1 Introduction In the EModE period, English underwent a number of substantial changes in all phonological subsystems. As a result, by the end of the period, English had evolved from the distinctly old-fashioned state of affairs preserved in Middle English to a system largely representing Present-day English. The language manifested shifts in the location of lexical stresses (see Section 2); it had cemented the phonological contrast between stressed and unstressed syllables (see Section 3); some distributional changes, a few new acquisitions as well as cluster reductions had yielded the present-day consonant system (see Section 4); and the vowel system had changed almost beyond recognition, now being characterized by an asymmetric relationship between long and short vowels and an almost complete renewal of the set of diphthongs (see Section 5). Yet further changes were to follow before the vowel system reached its present-day state. All of these aspects will in turn be highlighted in the following sections. Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 589–604
590
V Early Modern English However, what makes the descriptive task undertaken in this article easier than, for instance, a comprehensive outline of the phonology of Middle English, is the fact that Early Modern English was also a time of standardization: after an era of dispersion and dialectal diversity, the English language had regained most of the functions taken over by French in the centuries following the Norman Conquest. From the EModE era date numerous grammars, proposals for spelling reforms, rhyming dictionaries, shorthand guides and materials developed for teaching English, all of which provide useful sources for the reconstruction of EModE phonology (see the monumental work by Dobson 1968, drawing together much of this evidence). As a consequence of the generalized use of English for official purposes, the standardization of spelling, begun by the Chancery in the 15th century, continued at increasing speed, making written texts difficult to assign to any particular region. Of course, the standardization of written usage went considerably further than that of spoken dialects, which varied as a function of the social status of the speaker. In an often-quoted passage defining the origin of the newly risen English standard, Puttenham (1589) recommends the vsuall speach of the Court, and that of London and the shires lying about London within lx. myles, and not much aboue. I say not this but that in euery shyre there be gentlemen and others that speake but specially write as good Southerne as we of Middlesex or Surrey do, but not the common people of euery shire … (Puttenham 1589: 120–121).
Not surprisingly, it was thus the variety used in the political, cultural, and economic centre that set the norm to which other dialects were attracted, while differences persisted particularly in spoken usage. Possibly as part of this standardization process, the phonological makeup of many high-frequency words stabilized in one way or the other. While Middle English had been an era characterized by an unprecedented flexibility in terms of the presence or absence of variable segments, Early Modern English had lost these options. A wordfinal was no longer pronounceable as [ə]; vowel-final and consonant-final forms of the possessives my/mine, thy/thine, and of the negative no/none were increasingly limited to determiner vs. pronoun function, respectively; formerly omissible final consonants of the prepositions of, on, and in became obligatory, and the distribution of final /n/ in verbs was eventually settled (e.g. infinitive see vs. past participle seen). In ME times, this kind of variability had been exploited to optimize syllable contact at word boundaries by avoiding hiatuses and consonant clusters (e.g. my leg but min arm, i þe hous but in an hous, to see me but to seen it). The increasing fixation of word forms in Early Modern English came at the expense of phonotactic adaptability, but reduced the amount of allomorphy; in other words, phonological constraints were increasingly outweighed by morphological ones (cf. Schlu¨ter 2009b).
2 Word stress Stress assignment has since ME times been a hybrid of two conflicting systems of different origins. Throughout the history of English, native Germanic lexemes have as a rule had their stress on the first syllable from the beginning (unless they carried a stressless prefix). In contrast, Latin and French loanwords arrived with a stress system that counted syllables from the end and assigned stress to the first stressable (or heavy)
37 Early Modern English: Phonology syllable from there, with the final syllable as a rule being skipped in Latin. Predictably, Early Modern English, which contained and continued to adopt vast numbers of Romance loanwords, was characterized by a considerable vacillation with regard to the location of main accents. Application of the French stress rule led to accent on final syllables, e.g. pare´nt, prece´pt, colle´ague, where Present-day English has in many cases developed initial stress. Overall, educated speakers can be expected to have stuck to this rule longer than the less educated part of the population, who commonly generalized the initial stress rule, e.g. pa´rent, pre´cept, co´lleague. In some cases, the initial stress rule was extended to items where it did not eventually become established, giving pronunciations like co´nvenient, de´fective, pe´rspective, and su´ggestion (examples from Levins 1570). The aphesis of unstressed initial vowels attested in EModE loans like la´rum (< ala´rum) and spa´ragus (< aspa´ragus) seems likewise due to an overextension of Germanic initial stress. Alongside these Romance stress rules on the one hand and fully anglicized stress patterns on the other, the EModE system contained numerous exceptions. Thus, stressable final syllables and word-initial syllables in some cases remained unstressed, while wordmedial syllables received the main accent, e.g. demo´nstrate, emba´ssage, illu´strate, retı´nue, sono´rous. In the course of the later history of the language, the variability in terms of word accent that characterized the 16th and 17th centuries largely subsided in favor of stable patterns that were often a result of the history of individual words. Furthermore, for much of the Early Modern period, lexemes of three or more syllables tended to carry a relatively prominent secondary stress, separated by at least one syllable from the primary one (indicated here by grave accents). As a consequence, speakers tended to alternate between realizations like a´dvertı`ze and a`dvertı´ze, a´llego`rical and a`llego´rical, a´ccesso`ry and a`ccesso´ry, pa´ramo`unt and pa`ramo´unt, partı´cipa`te and partı`cipa´te. It is only in the 17th century that pronunciations that had lost the secondary stress began to compete with those that preserved it. In British English, the innovative variants ousted the older ones in subsequent centuries, while in American English secondary stress survives in many lexemes.
3 Syllable reduction Since the period under discussion, the presence or absence of stress has had important consequences for the phonological makeup of a syllable, in particular of its nuclear vowel. As early as Middle English, there had been a tendency for unstressed vowels to be reduced (which had gone to completion in final syllables), but the trend only gained considerable momentum in Early Modern English. At that time, vowels in non-primary stressed syllables could be of at least three kinds, corresponding to different degrees of reduction. In realizations that preserved secondary stress, long vowels and diphthongs were retained, e.g. pu´rpo`se /oː/, o´penly ` ~ /əɪ/, hı´story ` ~ /əɪ/, glo´rio`us /əʊ/, e´mpero`ur /əʊ/, ´ ` ´ ` certain /ɛɪ/, captain /ɛɪ/. Even when short and stressless, vowels remained clearly distinct in careful speaking styles, e.g. co´untenance /a ~ æ/, vı´llage /a ~ æ/, kı´ngdom /u/, se´ldom /u/, ca´ptain /e ~ ɛ/, o´penly /i/, e´mperour /u/. In less conscious styles, however, all vowels merged under one or two realizations, a centralized vowel /ə/ or /ʌ/, and a slightly higher /ɪ/, depending on the variety, the segmental context, and the lexeme. In appropriate contexts, unstressed vowels could be dropped completely (as is still the case in Present-day English): For one thing, in combinations with nasals or liquids, vowels could disappear in
591
592
V Early Modern English favor of a syllabic quality of the consonant, e.g. in bottom, garden, bottle, double, acre, and slender. For another, they could disappear in medial position of initially stressed words (which was in some cases indicated by the spelling), e.g. evry ~ every, sentry (< century), curtsy (< courtesy), fancy (< phantasy). (Note that these examples present a mixed set: While evry is a purely orthographical indication of syncope, the other three are instances of a beginning orthographic differentiation of a polysemous word, eventually resulting in a meaning split.) Phonological reduction up to and including the loss of unstressed vowels has since EModE times also characterized function words (prepositions, pronouns, auxiliaries etc.). Depending on factors such as speaking rate, formality, prosodic prominence, and syntactic independence, speakers had strong and weak forms at their disposal, e.g. and [and/ən], have [hɛːv/əv], would [wəʊld/wʊd], my [məɪ/mɪ], through [θrəʊx/θrʊ], you [jəʊ/jʊ]. Moreover, in high-frequency collocations, two function words could be contracted, leading to the loss of either the first (e.g. ’tis, ’twas, ’twill ) or the second vowel (e.g. they’re, we’ll, you’ve, can’t). While the phenomenon is not unknown in Old and Middle English (e.g. cham < ich am, het < he it, nabban < ne habban, nas < ne was), its currency increased considerably in the Early Modern period, and it is obvious from the examples given that the inventory of possible contractions underwent considerable change.
4 Consonants Compared to the vowel system, the consonant system remained fairly constant in the EModE period. The distribution of one or two consonants changed significantly, two phonemes were newly formed, and several initial and final clusters were simplified.
4.1 Distributional changes Initial /h/ had always been part of the English system, but had almost become extinct in ME times, with the exception of East Anglia and the Northeast. Thus, most early ME varieties spoken in England and Wales were completely /h/-less, not only in unstressed function words (where standard Present-day English still drops the /h/, e.g. he, his, him, her, has, have, had, etc.), but also in initial position of content words of Germanic origin (e.g. hand, heart, hair, house, husband, etc.). While the weakening of initial /h/ had been a natural continuation of its Old English demise in less prominent positions, and the arrival of French loanwords with mute (e.g. habit, hazard, heir, history, horror) played at most an auxiliary role, a reversal of the trend is traceable to later Middle English (from about 1350 onwards) and gained considerable momentum in the EModE era (cf. Schlu¨ter 2006, 2009a). The comeback of initial /h/ in Germanic as well as Romance words – spanning the 14th to 20th centuries – was certainly helped by the consistent preservation of ‹h› in the spelling. In large part, it was however a naturally occurring process whose functional motivation may be seen in the restitution of a consonantal syllable onset (which is universally preferred to vowel-initial syllables). The reinforcement of /h/ was retarded by factors such as the Romance origin of the lexeme, a high token frequency, and the absence of stress on its initial syllable, and it was speeded up by its native Germanic origin, a low frequency, and a primary stress on the first syllable (cf. Schlu¨ter 2009a). Throughout the Early Modern period, the
37 Early Modern English: Phonology realization of /h/ thus remained highly variable; witness numerous spellings like an hundred, myn husband, thyn humble servant, non history, etc., but it strengthened continuously long before /h/-dropping became stigmatized in the 18th century. In contrast, /h/’s allophones [c¸] (following front vowels) and [x] (following back vowels), which occurred in word-final position and before /t/, were lost in all southern English dialects in the course of the EModE era. The palatal variant [c¸] possibly disappeared somewhat earlier (15th to 17th centuries) than the velar one. It was vocalized and thus led to a compensatory lengthening of a preceding short vowel (e.g. high, night, thigh). Velar [x] was from the 14th century in some dialects replaced by the acoustically related consonant /f/, most typically when following /u/. For a while, /f/ had a wider spread in standard English than today, occurring for instance in daughter, bought, naught, taught, etc. By the mid-18th century, the modern distribution had been reached (cough, enough, laugh, rough, tough, trough, draught, and laughter). A distributional change that was to have important repercussions in the vowel system was the weakening of /r/ in non-prevocalic contexts. The process affected the southern British English standard and related dialects, but did not occur in Scotland, Ireland, most of the United States or Canada, which remained /r/-pronouncing (rhotic). It is assumed that the phonetic weakening of /r/ proceeded from an original trill or flap via an approximant stage to a complete loss of the consonantal closure. The first sporadic spellings testifying to /r/-loss occur in late Middle English of the 15th century, further evidence comes from EModE private writings, but the main changeover happened only in the 18th century, that is, after the end of the period focussed on here (cf. Lass 1999: 114–116). While Early Modern English was thus mostly rhotic, /r/-loss nevertheless deserves to be mentioned here because when following a stressed vowel, the weakening /r/ vocalized to produce a transitional /ə/ or led to a compensatory lengthening of the vowel as early as the 16th and 17th centuries. In addition, a following /r/ interfered significantly with the major reshuffling of the vowel system known as the Great Vowel Shift (see Section 5.2.1). Further consonantal changes concerned phonotactics, i.e. the ways in which consonants could be combined into consonant clusters. Cluster reduction, in particular, led to the creation of two new members of the consonant system, /ʒ/ and /ŋ/, which are therefore treated in Section 4.2.
4.2 Cluster reduction The reduction of onset clusters pursued the road already taken in Middle English. The consonant clusters /hr/, /hl/, /hn/, and /wl/ had shed their first members in early Middle English. The combination /hw/, with its first member weakening, continued to be distinguished from /w/ in general usage (though no longer in the southern dialects), so that which and witch, whine and wine were kept apart up to the 18th century (and continue to be in certain dialects, including Scottish and Irish English as well as a few American and Canadian accents). The reduction of the /wr/-cluster in items like write, wrong, and wrist began in the 15th and 16th centuries and was completed in the 17th. The clusters /gn/ in gnash, gnat, gnaw and /kn/ in knee, knit, know persisted somewhat longer, possibly in the form of assimilated /dn/ or /ŋn/ for /ɡn/ and /tn/ for /kn/. The simplification to /n/ began in the 17th century and was generally accepted in the South of England only in the early 18th century.
593
594
V Early Modern English An addition to the consonant inventory of Early Modern English resulted from an assimilation of the consonant sequence /zj/, occurring in medial position of French loan words like vision, occasion, and leisure. In the 16th century, the two components assimilated to form a new consonant /ʒ /. Comparable assimilatory simplifications affected the clusters /sj/, /tj/, and /dj/ in unstressed syllables. Rather than producing novel consonant phonemes, these added to the numbers of the existing /ʃ/-, /tʃ/-, and /dʒ/-phonemes, respectively. The 15th century saw the appearance of /ʃ/ in words like session, obligation, and mathematician, and the 17th century the rise of /tʃ/ in words like Christian, creature, and mutual and that of /dʒ/ in words like soldier, Indian, and grandeur. All of these new realizations, in particular the new /ʒ/-phoneme, took a long time to become accepted into the standard, and in some cases variation between consonant clusters and assimilated pronunciations persists to the present day. Further phonotactic changes concerned the reduction of final consonant clusters, which did not, however, become ubiquitous. Word-final /mb/ had already been reduced to /m/ in Middle English in items like bomb, dumb, lamb, plumb, tomb, etc. In the late 16th and 17th centuries, /ɡ/ was deleted after the velar allophone of /n/, making /ŋ/ phonemic. This loss occurred first in word-final position (e.g. sing, ring, strong, long) and then also in morpheme-final position (e.g. singer, ringing) except before adjectival inflections (e.g. stronger, longest). In the unstressed present participle ending -ing, /ɪŋ/ was further changed to /ɪn/ in many standard dialects. However, in later centuries, the /ɪŋ/-realization was enforced in standard usage and /ɪn/ restricted to rapid or colloquial speech. Finally, a minor consonantal change limited to a certain number of lexemes and dating to Early Modern English is the disappearance of /w/ when following another consonant and preceding a rounded back vowel, e.g. in sword, two, and who, and somewhat more systematically in unstressed syllables, e.g. Southwark, conquer, answer. In some further items, e.g. swollen, swoon, swore, awkward, boatswain, forward, housewife, and pennyworth, /w/ was later restored on the basis of the spelling or of related words. After the re-establishment of initial /h/, the loss of its allophones [c¸] and [x], and the introduction of /ʒ/ and /ŋ/, the EModE inventory of consonants was practically identical with the present-day one. All in all, the consonantal system of English has, however, remained relatively stable, in particular when compared to the fundamental turnover undergone by vowels in the same period of time.
5 Vowels The vowel system Early Modern English inherited from Middle English was very much unlike that of Present-day English. In EModE times, many changes happened that gave rise to a much more “modern” system. At the beginning of the period, the system was largely based on quantity contrasts that were inherited from Old and Middle English: many monophthongs occurred in pairs of long and short members, and the lengthenings and shortenings created by ME sound changes persisted, even within individual paradigms or word families, e.g. keep : kept, child : children, holy : holiday, wise : wisdom, wild : wilderness. The short vowels distinguished three and the long vowels four heights (both front and back). In addition, Middle English had a few closing diphthongs, whose second element was either /ɪ/ or /ʊ/. In the course of the Early Modern period, long and
37 Early Modern English: Phonology
595
short vowels developed in different ways, giving rise to quality differences in addition to the quantity differences. Many of the former diphthongs monophthongized, thus giving the English vowel system a less diphthongal character than it had either before or after Early Modern English. Some instances of the monophthongs, in particular ME /ɛː/ and /oː/ (corresponding to EModE /eː/ and /uː/), shortened, which accounts for those cases where the Present-day English spellings indicate length, but the vowels are pronounced short, e.g. bread, flood. In addition, many vowels underwent conditioned sound changes in specific phonological contexts, above all before /r/ and /l/. The most important changes will be discussed in what follows. The sections focus on unconditioned and conditioned changes, respectively, and treat long monophthongs, diphthongs, and short monophthongs in turn, but since these classes are to some extent interconnected through sound changes, this separation serves only expository purposes and will be deviated from in some places.
u
i υ
i e e
o
ui
iυ əi
ɔi
eυ əυ
ɔ
ε ε
aυ
a
ɔυ
ɒ
Figure 37.1: Early Modern English stressed monophthongs and diphthongs around 1600 (adapted from Go¨rlach 1994: 53)
Figure 37.1 depicts the location of monophthongs and diphthongs in the vowel chart around the middle of the EModE period. Some of the changes mentioned had already taken place at this time; some others were yet to come. Wavy lines in the diagram on the left show changes under way at the turn of the century; arrows in the diagram on the right indicate the starting points and targets of diphthongal realizations (but not the directions of diachronic changes).
5.1 Unconditioned changes The most momentous unconditioned change affecting all the long monophthongs in the Early Modern period is the Great Vowel Shift (GVS). The GVS raised ME /aː/, /ɛː/, /eː/, /ɔː/, and /oː/ by one step each, and the two ME vowels that were already maximally close, namely /iː/ and /uː/, diphthongized. Since the changes are assumed to be causally connected, the GVS represents a showcase example of a chain shift. While the initial signs of the shift go back at least to the 15th century, the first stage was completed in the 16th century. The results of this systematic shift can be seen in Figure 37.1, illustrating the status quo around the year 1600. Despite its systematic character and reasonably detailed documentation in contemporary orthoepic descriptions, the precise mechanics continue to be debated among linguists. Since the present volume contains a chapter expressly devoted to the GVS, this important part of EModE phonology will be omitted here (see Krug, Chapter 48).
596
V Early Modern English
5.1.1 Long vowels Depending on what is included under the concept of the GVS, the end point of the shift is also open to discussion (cf. Lass 1999: 80–85). After 1600, the long vowels continued to evolve, and the focus of the present discussion will be on these subsequent changes of the second half of the EModE period. While no mergers happened as part of the GVS chain shift, /iː/ (going back to ME /eː/) and /eː/ (going back to ME /ɛː/) fell together in /iː/ around the year 1700, though variation continued during the 18th century. This merger eliminated the opposition between items like see : sea and meet : meat. Another partial merger resulted from the further evolution of /ɛː/ (going back to ME /aː/) as in make, ale, ape, bake, drake, hate, knave, etc.: the sound raised to /eː/ in the 17th century and diphthongized to /eɪ/ around 1800. At the same time, not all instances of /eː/ (going back to ME /ɛː/) had in all dialects completed the rise to /iː/, so that they merged with items from the /ɛː/ class. In the 18th century, this merger became limited to a few lexemes in standard pronunciation, in particular grate : great and brake : break, while raised realizations came to prevail for other members of the /eː/ class, e.g. reach, leaf, clean (unless they had previously been shortened; see Section 5.2.1). ME /ɔː/ as in boat or no had, by virtue of the GVS, moved to /oː/. In the late 16th and 17th centuries, it was joined by ME /ɔu/ as in blow or know. Whether the vowel resulting from this merger was monophthongal or diphthongal is debatable and without doubt varied from one dialect to another (see Section 5.1.2). From the 18th century onwards, the diphthongal realization /oʊ/ prevailed in the standard (and became Received Pronunciation /əʊ/ around 1920). The ME monopthongs /iː/ and /uː/, which had diphthongized under the GVS, provide a transition to the field of the EModE diphthongs. The exact quality of the sounds resulting from the diphthongization is controversial. The versions with a centered initial element given in Figure 37.1, namely /əɪ/ and /əʊ/, represent the time-honored view still adhered to by most linguists, which has however been challenged by others – see, for instance, Faiß (1989) and Lass (1999), who reconstruct /ɛɪ/ and /ɔʊ/ (for further discussion, see Krug, Chapter 48). Be that as it may, in the second half of the EModE period, the new diphthongs continued to develop in a way to widen the distance between their first and second components. The modern situation, with /aɪ/ for the front diphthong and /aʊ/ for the back one was reached in standard British English in the 18th century.
5.1.2 Diphthongs Many of the diphthongs inherited from Middle English changed into monophthongs in the Early Modern era. The only two that remained were /ɔɪ/ and /uɪ/, which merged into /ɔɪ/ by the end of the period. The diphthong /ɔɪ/ occurred mostly in loan words from Old French such as choice, joy, noise, toy, while /uɪ/ was typical of the Anglo-Norman dialect of French that brought words like join, boil, coin, point, poison, and toil into Middle English. EModE texts show some variation between ‹oi› and ‹ui› forms (e.g. point/ puint, poison/puison), but the latter were less frequently used than the former and were eventually given up around the end of the 17th century. In a transitional phase, some members of the /uɪ/ group apparently merged with the reflexes of ME /iː/, when the latter reached the /əɪ ~ ʌɪ/ stage. Thus, loin : line and point : pint became
37 Early Modern English: Phonology homophones. This realization was however stigmatized as a provincialism and given up in favor of /ɔɪ/ in the 19th century. As already mentioned, all the other ME diphthongs are in standard accounts assumed to have turned into monophthongs in the EModE era, though in some cases, new diphthongs were established in the later evolution (cf. Barber 1997: 114– 115; Go¨rlach 1994: 55–56; Lass 1999: 91–94; Nevalainen 2006: 123–124). (Note, however, that there is no consensus among historical linguists concerning the evidence for the assumed monophthongizations of /ɔʊ/ to /ɔː/ and of /aɪ/ to /ɛː/ and their subsequent re-diphthongizations. It is probably more judicious to assume that the monophthongizations were only completed in some dialects, while others preserved more or less markedly diphthongal realizations and played a role in their establishment in the Late Modern English standard.) To start with, /aʊ/ became /ɒː/ in the mid-17th century, thus filling the gap in the open back corner of the long vowel system (Figure 37.1), and later changed into /ɔː/, e.g. in cause, law, and taught. Second, /ɔʊ/ monophthongized in the second half of the 16th and first half of the 17th century to /ɔː/ and took part in the GVS-conditioned raising to /oː/ (which became /oʊ/ in Late Modern English), e.g. in bowl, flow, know, low, and soul. Third, the ME /aɪ/ diphthong narrowed to /ɛɪ/ and subsequently monophthongized to /ɛː/, which joined the GVS-raising to /eː/. As a result of this merger with former /aː/ as in make, a large number of homophones were created in the mid-17th century, e.g. days : daze, hail : hale, raise : raze, tail : tale. In Late Modern English, the monophthong developed into a new diphthong /eɪ/, representing the modern state of affairs. In addition, Middle English had two diphthongs, /eʊ/ (occurring in words like beauty, dew, few, hew, and newt) and /iʊ/ (occurring in words like chew, due, hue, new, and true), that eventually collapsed under /juː/. Phonetically, the change proceeded by a raising of the first component of /eʊ/ to /i/ and a shift of the vocalic centre of the diphthong from the first to the second component. As a result, /i/ was reanalyzed as the glide /j/ and assigned to the onset, while /ʊ/ lengthened to compensate for the loss of vowel quantity. The chronology of the merger of /eʊ/ and /iʊ/ on the one hand and the development from diphthong to glide plus long monophthong on the other is somewhat disputed. In Lass’s (1999: 99–100) account, the merger and the development of /juː/ coincided in the second half of the 17th century. In contrast, Dobson’s evidence indicates that the development of the glide began as early as the late 16th century, while the loss of the distinction between /eʊ/ and /iʊ/ proceeded slowly south- and westwards and was completed in the standard by the last third of the 17th century. According to this chronology, the realization /juː/ must have appeared earlier for the /iʊ/ words than for the /eʊ/ words, but the difference should have been neutralized by the late 17th century (cf. Dobson 1968: 700–713, 798–799). On the basis of the corpus study provided in Schlu¨ter (2006), both datings of the relevant changes stand in need of revision. For one thing, the emergence of the glide in the reflex of ME /iʊ/ can be traced back at least to the late 16th century (confirming Dobson thus far). For another, former /eʊ/ words show the first signs of an emerging glide only in the late 18th century and keep lagging behind /iʊ/ words as late as the end of the 19th century. This suggests, pace Lass and Dobson, that the merger was not completed until the 20th century. Furthermore, it indicates that while the evolution of the /j/ onset in /iʊ/ words is properly part of an account of EModE phonology, the merger of (former) /iʊ/ and /eʊ/ as well as the adoption of /j/ by the latter belongs to later chapters of the history of English. The
597
598
V Early Modern English same is true of the dropping of /j/ that later reduced /juː/ to /uː/ in certain environments. The reduction occurred early after /r/, /dʒ/, and /tʃ/, as in rude, June, and chew, at the beginning of the 18th century extended to /l/ and /s/ as in clue and suit, and in American English further to /t/, d/, and /n/ as in tube, due, and new.
5.1.3 Short vowels Turning to the set of short vowels, the spontaneous, unconditioned changes that occurred in the EModE era were much less dramatic than the GVS or the largescale reduction of diphthongs to monophthongs. Among the three short front vowels, we find a moderate degree of variation with regard to height. There is some disagreement in the literature on whether the close vowel /i/ as in bit, thin, and give was the first or the last one to move. Many authors hold that it lowered and centralized to /ɪ/ as early as Middle English (Faiß 1989: 33–34; Nevalainen 2006: 124; Stockwell and Minkova 1990, 2002), while Lass (1999: 88) concludes from contemporary descriptions that the change occurred only towards the end of the Early Modern era. In some dialects, including London, /ɪ/ lowered even further, leading to spellings like menysters ‘ministers’ and cete ‘city’, but these variants were not adopted into the standard. Similarly, the precise quality of the vowel of bed, set, and rest, occupying the middle height, is somewhat unclear. Opinions diverge on whether the ME vowel was /e/, which lowered to /ɛ/, or conversely, /ɛ/, which raised to /e/ in the first half of the EModE period (cf. Lass 1999: 87 vs. Barber 1997: 109 and Nevalainen 2006: 124, respectively). Concerning the lowest among the ME short vowels, instantiated in bad, man, and rap, linguists disagree about its history before the present-day state of /æ/ was reached in the mid-17th century. The widespread view according to which Middle English /a/ raised to /æ/ in the early 17th century (Faiß 1989: 36; Lass 1999: 85; Nevalainen 2006: 124) is contested by Minkova (2001: 85), who concludes from scribal and rhyme evidence that a higher /æ/ vowel survived from Old English in southern dialects of Middle English. Both of the short back vowels became more open in Early Modern English. The midlow /ɔ/ as in dog, hot, and rob lowered further to /ɒ/, which was well established by 1670. Occasionally, it merged with the more front /a ~ æ/ in fashionable realizations of the second half of the 17th century (cf. contemporary spellings like plat ‘plot’ and Gad ‘God’). In most Northern American dialects, /ɒ/ pursued a different path: it lengthened and unrounded to /ɑː/ in the late 17th or early 18th century. Finally, as in the case of /i/ > /ɪ/, the change of the high back vowel from /u/ to /ʊ/, which is generally dated to (late) Middle English (Faiß 1989: 39; Stockwell and Minkova 1990), is argued by Lass (1999: 88) to have occurred only towards the end of the 17th century. As a rule /ʊ/ lowered and unrounded to /ʌ/ not long after the year 1600 in the standard dialect (though not in the Midlands and the North of England). In certain conservative environments, however, it remained as /ʊ/. (The specific environments concerned and the resulting phonemic split will be discussed at the end of Section 5.2.2.) As is obvious from the above discussion, it is virtually impossible to assign to the short vowels of Early Modern English a precise location in the vowel chart. While it can be considered certain that short vowels were more lax than long vowels at the end of the period, it is unclear if and to what extent the same was already true at its beginning. The many controversies revolving around the issue are doubtless due to the fact that by their very nature, short vowels tend to have less extreme realizations
37 Early Modern English: Phonology than long ones and are therefore harder to describe, a problem that is exacerbated by the fact that the short-long pairings of vowels, characteristic of earlier stages of the language, were lost by virtue of the outcomes of the GVS.
5.2 Conditioned changes In addition to the general trends outlined in Section 5.1, both long and short vowels were implicated in a large number of conditioned sound changes. The present section concentrates on the most important sound changes conditioned by specific phonological environments.
5.2.1 Long vowels While unconditioned changes generally left vowel quantities intact, conditioned (or combinative) changes could involve quantities, qualities, or both. In many cases where the modern spelling with a double vowel grapheme indicates a long vowel but the vowel is pronounced short, it was shortened after 1400, which is about the time when English spelling conventions were fixed. In the EModE era, shortenings concerned, above all, the vowels /eː/ and /uː/ and occurred in monosyllabic words ending in /d/, /t/, /θ/, /f/, /k/, or /v/. Examples of the reduction of /eː/ to /e ~ ɛ/ include bread, dead, lead (N), sweat, death, breath, and deaf; examples illustrating the reduction of /uː/ to /u ~ ʊ/ are good, stood, hood, foot, book, took, and look. The reduction process was not a monolithic one, but reached individual lexemes at different points in time. This becomes evident when it is seen against the background of the unconditioned change from /u/ to /ʊ/, which has been dated in Section 5.1.3 to a time shortly after the turn of the 17th century: Consequently, the items flood, blood, and glove must have been shortened by the early 17th century to take part in the change to /ʊ/; the items mentioned above (good, stood, hood, etc.) were shortened later than that and therefore failed to undergo lowering and unrounding; finally, items retaining long /uː/, like mood, food, rood, and shoot resisted the reduction to /u ~ ʊ/. As mentioned in Section 4.1 above, the presence of a following /r/ often interfered with the regular evolution of long vowels under the GVS. While rhoticity was largely preserved throughout the EModE period, the /r/ had a variety of effects on the preceding vowel. First, as early as the 15th century an additional [ə] appears to have been inserted between a vowel and the /r/. This has eventually resulted in the modern centring diphthongs, e.g. /ɪə/ in dear, /ɛə/ in bear, and /ʊə/ in poor. Second, due to the complex articulatory movements involved in the production of /r/ – Lass (1999: 108) describes it as an alveolar or post-alveolar approximant with a velar plus a pharyngeal secondary articulation – it often exerted a lowering and rounding influence on the preceding vowel, but in some cases it also had a raising effect. The lowering effect of postvocalic /r/ in particular had the potential to counteract the raising of vowels under the GVS. Developments in the rows of front and back vowels will be treated in turn here since shifts in vowel height often led to (partial) mergers within these rows. For instance, /eːr/ was partly lowered to /ɛːr/ in the 15th and 16th centuries. Therefore, instead of the expected spellings deer and heer for the reflexes of OE de¯ore and he¯ran, we find dear and hear today, bearing witness to the lowered realization. In the examples given, the normal development to /iːr/ and later /ɪə/ has eventually
599
600
V Early Modern English prevailed. In contrast, lexemes containing /ɛːr/ did not in general undergo raising under the GVS and thus did not merge with the former class. They preserve unraised /ɛə/ to the present day, e.g. bear, pear, swear, wear. There are however a few lexeme-specific exceptions to this rule, including shear, spear, fear, ear, whose vowels did raise and merge with /eːr/ in Early Modern English, giving PDE /ɪə/. The /ɛːr/ (> PDE /ɛə/) group was partly joined by the reflexes of ME /aːr/, which closed to /æːr/ and /ɛːr/ under the GVS, involving examples like bare, fare, hare, pare, and share. Thus, while ME /ɛː/ as a rule merged with higher /eː/ (the merger of sea and see; cf. Section 5.1.1), before /r/ it collided with the next lower ME vowel /aː/. The changes undergone by the row of back vowels under the influence of a following /r/ were even more unpredictable. Starting from the lowest member, ME /ɔːr/ raised temporarily, approximately between 1650 and 1750, to /oːr/, only to lower again to /ɔːr/, where it stayed, evolving into PDE /ɔə/ around 1800, e.g. oar, lore, and more. At the same time, some instances of ME /oːr/ raised to and merged with /uːr/ around the year 1650, giving PDE /ʊə/, as in poor and moor. Most members of both the raised and the unraised groups later, around the year 1700, lowered to and merged with /ɔːr/, thus also ending up as /ɔə/, e.g. door, floor, whore and the alternative pronunciation of poor. Finally, ME /uːr/ developed regularly to /aʊr/ and then /aʊə/, as in flower, shower, and our. However, when followed by another consonant, it lowered and merged with /oːr/ and then /ɔːr/ as in court, mourn, and source. A last combinative change interfering with the regular development of a long vowel concerns ME long /uː/. After the glides /w/ and /j/ and in the context of labial consonants, it did not diphthongize, as would have been normal in the Early Modern period, but remained /uː/ instead. Examples include you, youth, wound, swoon, room, stoop, droop, loop, and tomb. This arrest of the regular development can be clearly attributed to functional constraints opposing a dissimilation of articulatory gestures.
5.2.2 Short vowels This section surveys conditioned changes in the area of short vowels, which again involved quantitative as well as qualitative changes that were frequently interconnected. What is more, in a few cases the changes led to the creation of new phonemes out of conditioned allophones, in particular when the conditioning environment was lost or one or the other set of allophones was augmented by additional members as a result of an independent sound change. Once again, a following /r/ played an important role as a factor interfering with regular developments. For instance, the short vowels /ɪ/, /ʊ/, and /e ~ ɛ/ became indistinct before an /r/ in final position or followed by another consonant, e.g. in bird, firm, sir, murder, hurt, and curb. The three vowels collapsed under a mid-central /ʌ/ or /ə/, which after the loss of rhoticity evolved into /ɜː/, and thus added a new phoneme to the set of long vowels. This change first affected /ɪ/ and /ʊ/ and only later reached /e ~ ɛ/; it started in the North and East in the 16th century, reached London by the 17th century and was complete only by the turn of the 19th century. However, in some items, /e ~ ɛ/ had been lowered to /a/ in the 15th century and thereby escaped this change (if only temporarily). Typical spellings from the long period of vacillation that indicate this realization include clark, dark, far, harvest, heart, starre, saruant, sarvice, and marcy. As can be inferred from these examples, the lowered realization was
37 Early Modern English: Phonology generally retained in Germanic words, whereas /e ~ ɛ/ was reintroduced (along with the elimination of spelling variants with ‹a›) in most loan words, where it eventually merged with /ɜː/. The remaining two short vowels, /a ~ æ/ and /ɔ ~ ɒ/ did not change to the same extent. Their lower allophones were favored in the context of a following /r/ and were lengthened to EModE /aː/ and /ɒː/, respectively. Around 1800, the front /aːr/ changed into /aːə/, which was then reduced to /ɑː/, as in arm, bar, cart, garden, harm, mark, and sharp. Similarly, the back /ɒːr/ evolved into /ɒːə/ and was reduced to /ɔː/ at the beginning of the 19th century, e.g. border, corn, for, horse, and north. Apart from /r/, the other liquid /l/ in the English consonantal system had comparable, though less pervasive effects on preceding vowels, in particular on /a ~ æ/ and /ɔ ~ ɒ/. When /l/ was followed by a word boundary or another consonant, an additional /ʊ/ glide was inserted between it and the preceding /a/ or /ɒ/ in the 16th century, resulting in the formation of the closing diphthongs /aʊ/ and /ɒʊ/. Where the /l/ was followed by another consonant, it was in addition totally assimilated to the /ʊ/; in other words, it vocalized. Subsequently, the new diphthongs were variably reduced to monophthongs, alternated with them throughout the 17th century and were ousted by them after the end of the Early Modern period. More precisely, /aʊ/ became /aː/ when it preceded (assimilated /l/ and) a labial consonant as in alms, balm, calf, calm, half, and palm, and it became /ɔː/ elsewhere, e.g. in all, ball, call, balk, chalk, stalk, talk, and walk. On the other hand, /ɒʊ/ became /oː/, which joined the regular post-GVS diphthongization, turning into /oʊ/ and later /əʊ/, as in roll, toll, colt, folk, holm, and yolk. A further combinative sound change concerned the evolution of /a ~ æ/ after /w/. Where /a/ followed /w/ and did not precede a velar plosive (/k/ or /ɡ/), it did not raise to /æ/, but backed and rounded to /ɔ/ in the course of the 17th century, e.g. in what, warm, wand, but not in whack and wag. Like the lack of diphthongization of long /uː/, mentioned above, the velarizing effect of /w/ is explained by the velar articulation of the glide itself. Turning now to some further changes conditioned by specific phonological contexts, we witness the creation of two new vowel phonemes, one short and one long. First, as already pointed out in Section 5.1.3, short /ʊ/ lowered and unrounded to /ʌ/ in an unconditioned change taking place at the beginning of the 17th century. However, /ʊ/ remained in many cases after labial consonants (/p/, /b/, /f/, and /w/) and before /l/ or /ʃ/, though this arrest of the change did not apply across the board. For instance, /ʊ/ was preserved in bull, full, bush, put, and wolf, but /ʌ/ established itself in but, buff and fuss. In addition, as mentioned above, lexemes containing shortened /uː/ increased the numbers of both /ʊ/ (e.g. good, stood, hood, etc.) and /ʌ/ (e.g. flood, blood and glove). As a result, the former short /ʊ/ phoneme split in two, namely /ʊ/ and /ʌ/, contrasting, for instance, in look and luck. Second, the short /a ~ æ/ was lengthened in the South in the 17th century where it preceded one of the voiceless fricatives /s/, /f/, or /θ/ (but not /ʃ/) or the clusters /ns/ or /nt/, for instance in glass, pass, castle, last, chaff, staff, bath, path, dance, and plant. This produced a new low front phoneme /aː ~ æː/ in a slot that had been vacated as a result of the GVS. The lengthened realization was stigmatized to some extent, so that it retracted in some lexemes, but became established in others. In 18th-century British English, the novel phoneme backed to /ɑː/, where it was later joined by instances of /a ~ æ/ that had lengthened before /r/ (arm, bar, cart, garden, etc., discussed above).
601
602
V Early Modern English This change did not happen in American English, which preserves /æː/ to the present day. A related change affected short /ɔ ~ ɒ/, which was lengthened at the same time and in the same environments as /a ~ æ/, i.e. before voiceless fricatives. Examples include loss, off, and cloth. Rather than creating a new phoneme, in the late 17th century the sound merged with /ɒː/ (which resulted from the monophthongization of ME /aʊ/ in the middle of that century). For a while, long and short versions coexisted side by side, but again the long ones were partly stigmatized. In contrast to the lengthened /a ~ æ/, short /ɒ/ was eventually restored before voiceless fricatives (except for some speakers of southern dialects). Summing up Section 5.2, it turns out that, compared to the set of long vowels, short vowels were considerably more liable to conditioned changes. While phonotactic contexts mainly interfered with GVS-related raisings in long vowels, their influence on short vowels did not remain as limited: they led to important qualitative and quantitative changes, a large-scale merger before /r/ and a phonemicization of two (and after the EModE era, three) allophonic contrasts, namely /u ~ ʊ/ vs. /ʌ/ and /a ~ æ/ vs. /aː ~ æː/ (and later /ɪ/, /ʊ/, and /e ~ ɛ/ vs. /ɜː/). The functional reason for the greater liability of short vowels to conditioned changes can be found in the fact that their articulatory targets are not as clearly defined as those of long vowels.
6 Summary To sum up, Early Modern English has revealed itself to be a period of massive changes, even discounting those subsumed under the Great Vowel Shift (Krug, Chapter 48). For one thing, the considerable variability in the domain of lexical stress patterns was limited in favor of one or the other pattern along the lines of either Germanic or Romance stress rules. For another, syllables not carrying the word accent were increasingly reduced, giving English the stress-timed rhythm that characterizes it today. And for another, the inventory of consonants acquired its present structure, albeit without undergoing any massive changes. The loss of the palatal and velar allophones of /h/, [c¸] and [x], is counterbalanced by the establishment of the phonemes /ʒ/ and /ŋ/. In addition to many consonant clusters given up in Middle English, a few further combinations were simplified by dropping their first or second member, in particular /wr/, /ɡn/, /kn/, and /ŋɡ/, or by assimilating both into a single consonant, as in the case of /zj/, /sj/, /tj/, and /dj/. The area that underwent the most dramatic changes in EModE phonology were the vowels. Among the short vowels, there was a limited amount of variation and change. But even beyond the GVS, the long monophthongs continued to change, which gave rise to noticeable qualitative differences between short and long vowels: both /eː/ and /ɛː/ raised further, in the first case leading to the see : sea merger, and the trajectory of the newly formed diphthongs /əɪ/ and /əʊ/ (or /ɛɪ/ and /ɔʊ/) widened to /aɪ/ and /aʊ/. Most of the ME diphthongs monophthongized, with the exception of /ɔɪ/ (< /ɔɪ/ and /uɪ/). Last but not least, an important number of conditioned vowel changes took place in the Early Modern era, involving certain shortenings and lengthenings in addition to qualitative changes (or the arrest of such changes). The contexts responsible for these combinative changes were, above all, a following weakening /r/, but also a following /l/, a preceding /w/, a preceding labial consonant, or a following voiceless fricative, plus some further segmental environments. What were allophonic differences at the
37 Early Modern English: Phonology
603
outset evolved into novel phonemes in the cases of /aː ~ æː/ and /ʌ/ (and, after the loss of non-prevocalic /r/, also /ɜː/).
u υ
i i e e
o ε
ɔi
aυ
ai
a
ɒ ɒ
Figure 37.2: Early Modern English stressed monophthongs and diphthongs around 1700
Despite these massive changes, the vowel system at the end of the period, represented in Figure 37.2, was still markedly different from the Present-day English one. It remained for later periods to develop the strongly diphthongal ring that characterizes English vowels today.
7 References Barber, Charles. 1997. Early Modern English. 2nd edn. Edinburgh: Edinburgh University Press. Dobson, Eric J. 1968. English Pronunciation 1500–1700. 2 vols. 2nd edn. Oxford: Clarendon. Early English Books Online. 1999. Ann Arbor: Proquest Information and Learning. http://eebo. chadwyck.com/home/ Faiß, Klaus. 1989. Englische Sprachgeschichte. Tu¨bingen: Francke. Go¨rlach, Manfred. 1994. Einfu¨hrung ins Fru¨hneuenglische. 2nd edn. Heidelberg: Winter. Lass, Roger. 1999. Phonology and morphology. In: Roger Lass (ed.), The Cambridge History of the English Language. Vol. III. 1476–1776, 56–186. Cambridge: Cambridge University Press. Levins, Peter. 1570. Manipulus Vocabulorum. London: Henrie Bynneman. (Source: Early English Books Online.) Minkova, Donka. 2001. Review of Lass, Roger (ed.). 1999. The Cambridge History of the English Language Vol. III. 1476–1776. Journal of English Linguistics 29: 83–92. Nevalainen, Terttu. 2006. An Introduction to Early Modern English. Edinburgh: Edinburgh University Press. Puttenham, George. 1589. The Arte of English Poesie. London: Richard Field. (Source: Early English Books Online.) Schlu¨ter, Julia. 2006. A small word of great interest: The allomorphy of the indefinite article as a diagnostic of sound change from the sixteenth to nineteenth centuries. In: Nikolaus Ritt, Herbert Schendl, Christiane Dalton-Puffer, and Dieter Kastovsky (eds.), Medieval English and its Heritage: Structure, Meaning and Mechanisms of Change, 37–59. Frankfurt: Lang. Schlu¨ter, Julia. 2009a. Consonant or “vowel”? A diachronic study of initial ‹h› from early Middle English to nineteenth-century English. In: Donka Minkova (ed.), Phonological Weakness in English: From Old to Present-Day English, 168–196. Houndsmills, Basingstoke/Hampshire/ New York: Palgrave Macmillan.
604
V Early Modern English Schlu¨ter, Julia. 2009b. Weak segments and syllable structure in Middle English. In: Donka Minkova (ed.), Phonological Weakness in English: From Old to Present-Day English, 199–236. Houndsmills, Basingstoke/Hampshire/New York: Palgrave Macmillan. Stockwell, Robert and Donka Minkova. 1990. Early Modern English vowels, more o’Lass. Diachronica 7: 199–215. Stockwell, Robert and Donka Minkova. 2002. Interpreting the Old and Middle English close vowels. Language Sciences 24: 447–457.
Julia Schlu¨ter, Bamberg (Germany)
38 Early Modern English: Morphology 1 2 3 4
Nominal inflectional morphology Verb morphology Derivational morphology References
Abstract By the end of the Middle English period there is already considerable loss of inflectional morphology, and in Early Modern English we see the last reflexes of a shift from synthetic Old English to analytic Modern English (Lass 1999: 139). In fact, the inflectional system of Early Modern English is not very different from what we have today (Go¨rlach 1991: 79). The changes in inflection which do take place between 1500 and 1700 show marked sociolinguistic differentiation and are the subject of well-known case studies in sociohistorical linguistics. The derivational morphology of Early Modern English, on the other hand, is considered to demonstrate much more wholesale and radical change in the form of new Latin prefixes and suffixes reanalyzed from borrowed lexis. The rate of integration of these word-formation processes is not, however, very uniform, and capturing this diversity is a major aim of this survey.
1 Nominal inflectional morphology 1.1 Nouns Gender marking on nouns was already lost by late Middle English. The only case marking left by 1500 is the genitive -s with the same allomorphs (/ɪz/, /s/, and /z/) as the plural morpheme (Barber 1997: 145). The use of the apostrophe s (’s) for the spelling of the possessive singular is not common until the late 17th century, and the s apostrophe (s’) for the possessive plural is not common until the late 18th century (Barber 1997: 143; Go¨rlach 1991: 82). The analytic variant, the of genitive, is available from late Middle English but becomes markedly more popular over the Early Modern period. The -s genitive tends to occur with human nouns and on modifiers in subjective relation to the head (the boy’s arrival) and the of genitive tends to occur with inanimate nouns and on modifiers in objective relation to the head (the release of the boy). This pattern remains quite Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 604–620
38 Early Modern English: Morphology consistent in the 17th century. At this time the -s genitive is regarded as somewhat more informal (Altenberg 1982; Rissanen 1999: 201–202). A much discussed construction associated with Early Modern English is the “his genitive” (the Kinge his fool). This is widespread in the 16th and 17th centuries, but in fact arose earlier (12th century) due to the homophony of the genitive morpheme and weak forms of his with /h/ deletion (Barber 1997: 146; Lass 1999: 146). It may have been a popular feature which then in the 16th century made its way into “respectable” prose (Go¨rlach 1991: 81). An oft-cited example from Shakespeare is the Count his gallies (Twelfth Night). The construction was extended to feminines in the 16th century, as in Lyly’s Juno hir bed, and apparently to plurals, as in the vtopians their creditors (Robinson’s translation of More’s Utopia 1551). However Allen (2008) has shown some well-known examples such as the latter, which is cited in the OED, to be misanalyzed cases of apposition. Typically the construction is restricted to proper names ending in sibilants which would otherwise have no formal marker of possession as in Glanvill’s Democritus his Well and Hercules his Pillars (Barber 1997: 146; Go¨rlach 1991: 81). Number marking with inflectional -s is highly regularized in Early Modern English. In Middle English the unstressed schwa of [əz] was lost except after sibilants, and this was followed by assimilation to preceding voiceless consonants, giving three allomorphs /ɪz/, /s/, and /z/. This allomorphy is more or less established by the 15th century, but unexpected forms in Hart’s transcriptions of 1569 such as birds, prinses, and faultz show that the system is not stabilized until about 1600 (Barber 1997: 144; Lass 1999: 141–142). Some of the mass nouns of Modern English are count nouns in Early Modern English (salmons, trouts). Conversely some nouns that today have an -s plural today could take a zero plural in Early Modern English (board, brick). Horse, winter, year, and lamb in Early Modern English are variable. Umlaut plurals (mice, geese) are in decline by Middle English and the older Old English plural in -en (as in oxen and children) is used only for deliberate archaism such as Spenser’s eyen, foen, skyen (Lass 1999: 141; Barber 1997: 145; Go¨rlach 1991: 80).
1.2 Pronouns Unlike nouns, pronouns in Early Modern English are still marked for person and gender as well as number and case. The EModE paradigm in Table 38.1 shows that as per the ME development, gender is marked in the third person only. Although the /h-/ of neuter hit was lost in Middle English, some claim that hit was still in use in the 16th century (Barber 1997: 150). In the late 16th century, its emerges as the neuter possessive pronoun, replacing his (Lass 1999: 148; Go¨rlach 1991: 85–86). His can still be observed in the Authorized Version, as in (1): (1)
if the salt haue lost his savour, wherewith shall it be salted? (1611 King James Bible Matthew 5:13; Barber 1997: 150)
Mine and thine as determiners are common before vowels and /h/ in the 16th century, but by the 17th century attributive /-n/ forms are rare (Barber 1997: 152; Go¨rlach 1991: 85). Changes in the neuter third person pronoun and the system of second person pronouns are shown with arrows.
605
606
V Early Modern English Table 38.1: Early Modern English personal pronouns (Nevalainen 2006: 77) Person/ number
Subjective
Objective
Possessive determiner
Possessive
1P SG 1P PL 2P SG 2P PL 3P SG personal 3P SG non-personal 3P PL
I we thou ~ ye → you ye → you he, she (h)it → it they
me us thee ~ you you him, her him, (h)it → it them
my/mine our thy/thine → thy ~ your your his, her his → its (of it) their
mine ours thine ~ yours yours his, hers (his >its) theirs
One of the most remarkable developments in the pronoun system of Early Modern English is the emergence and then decline of social deixis in the second person. You, historically the plural form, became used in Middle English, under courtly French influence, as a polite or deferential singular (Barber 1997: 153; Go¨rlach 1991: 85). In a parallel change, nominative ye ceases to be an alternative to you (complete by 1600) and you becomes the form for the nominative and the accusative. Yet English did not develop a typically European T/V system (Brown and Gilman 1960) with reciprocal thou (T) encoding intimacy and solidarity and non-reciprocal T/V encoding asymmetry in power or status (Brown and Gilman 1989: 177; Lass 1999: 149; Wales 1983). In the middle of the Early Modern period, you is the polite form used by inferiors to superiors, but it is also a neutral and unmarked form among the upper classes. The general use of you spread down the social hierarchy and “by 1600, you was the normal unmarked form for all speakers with any pretension to politeness” (Barber 1997: 155). Thou was retained to occasionally mark asymmetrical relationships; mostly it had an “emotional” use to convey intimacy and affection, sometimes contempt. These affective shifts are reflected in the switching of pronouns by the same interlocutors even within the same text. Some evidence comes from dramatic dialogue: In Macbeth, Malcolm addresses Macduff with you, a proper form for a Scottish thane, until Malcolm’s emotional statement “but God above deal between me and thee” (IV.iii.120–121; Brown and Gilman 1989: 177). There are also abundant examples from private letters. Sir Thomas More, who otherwise addresses his daughter as you, says “Surely Megge a fainter hearte than thy fraile father hath, canst you not haue” with the concord for thou applied to you (Lass 1999: 151). Thou becomes increasingly restricted to high registers by the end of the 17th century, although it is also associated with regional use (Nevalainen 2006: 18) (see Busse, Chapter 46).
1.3 Adjectival comparison In Early Modern English the only morphological marking on adjectives is the comparative and superlative degrees of comparison (-er, -est). The periphrastic expression of gradation (more, most) had already become common in Middle English, providing two systems. In the modern system periphrasis is in complementary distribution with suffixes: monosyllabic bases take suffixes (bigger, biggest), disyllables prefer suffixes, but can take periphrasis (hairier, more hairy); trisyllabic and longer forms take
38 Early Modern English: Morphology periphrasis (*beautiful-er). This situation is not completely established in Early Modern English, however. We find forms like easilier and more brief in John Hart’s Orthographie of 1569, famousest and difficultest in Milton, learneder in Johnson, and ragingest in Nash. Double comparison was more common in the 16th and 17th centuries, illustrated by Shakespeare’s “this was the most unkindest cut of all” (Julius Caesar) and “more nearer” (Hamlet). There is also apparently more free variation: Ben Jonson uses both fitter and more fit, Shakespeare uses sweeter and more sweet (Lass 1999: 156–158; Barber 1997: 136–147). Go¨rlach (1991: 83–84) believes that the periphrastic form was more associated with written or educated language and that much of the loss of the inflected form for disyllabics was due to prescriptivism. However, studies of the Helsinki and ARCHER corpora (Kyto¨ 1996; Kyto¨ and Romaine 1997) suggest that the inflectional forms reassert themselves after 1700.
2 Verb morphology 2.1 Person and number The second person continues to be marked in Early Modern English in concord with the pronoun thou, but falls into disuse along with thou in the 17th century (Barber 1997: 164–165; Go¨rlach 1991: 88; Lass 1999: 139). The second person marker -st appears on the present (bearest, giuest, walkest) and the past (barest, gauest, walkedst). Third person plural is marked in the present by the Midlands variant -en in 15th century texts, as in (2): (2)
Southern western & northern men speken frenssh all lyke in soune & speche (1480 The Description of Britain [Caxton edn.]; Go¨rlach 1991: 89)
The marker falls away quickly in the 17th century from the standard language. The normal plural for Early Modern English is the uninflected form (Barber 1997: 170–171). Although there is only one marker of third person singular in Modern English, -s is in competition with -eth throughout the Early Modern English period. The -s form was originally northern and had spread to the East Midland system by the 15th century. The original southern -eth form became the standard written form when the new standard literary language took shape. Yet -s continued to move southwards and in 1500 was probably common in southern speech. The use of -s increases and over the 16th century it becomes the normal spoken form (Barber 1997: 166–167; Lass 1999: 162–164; Nevalainen 2006: 17). More precisely, variation in the early stages is between -eth and -es (as in comyth and makys) rather than the contracted -s and the syllabic -eth which we find in the 17th century (Nevalainen and Raumolin-Brunberg 2003: 67–68). Yet it would be simplistic to think in terms of a spoken variant and a written variant. Rather, -eth is associated with more formal text types, namely official documents, poetry, sermons, and biblical translations (such as the Authorized Version of 1611); and -s appears in journalistic prose, drama, private letters, and diaries (Barber 1997: 166–168; Go¨rlach 1991: 88; Nevalainen and Raumolin-Brunberg 2003: 81). Studies of the variation in Shakespeare’s plays reveal rapid change over a short critical period (Taylor 1976; Stein 1988). More longitudinal research using the Corpus of Early English
607
608
V Early Modern English Correspondence (CEEC) shows two waves of change. In the “first wave” in the latter half of the 15th century, the change to -s is led socially by the “lowest literate ranks”. In the second wave, around 1600, the middle or upwardly mobile ranks lead this change, especially women in these ranks (Nevalainen and Raumolin-Brunberg 2003: 121–122, 140, 144, 178–179). The (present) inflected forms of HAVE and DO (hath and doth) retain the older form for much longer, but it may be that these continued to be used as written forms after the spoken use of has and does (Lass 1999: 163–165; Barber 1997: 168). Modal auxiliaries were normally not inflected for the third person singular (unless they are also still lexical verbs as in he dares and he willeth), but they do have the second person singular inflection (thou canst). The second person singular forms of shall and will are shalt and wilt (Barber 1997; Go¨rlach 1991: 89).
2.2 Tense, mood, and aspect All weak verbs in Early Modern English as today are marked in the past tense. The Modern English system of allomorphy of the past tense marker -ed was not established until 1600: /əd/ after /t/ or /d/ (waited, heeded), /d/ after a vowel or voiced consonant (died, begged) and /t/ after a voiceless consonant (looked, wished) (Barber 1997: 174). There was considerable variation into the 18th century (Lass 1999: 172), and the /əd/ pronunciation with the schwa vowel, which began to be lost in the 16th century (Go¨rlach 1991: 92) could be used in more positions than is possible today. Syncope is indicated around 1600 by spelling (begd, lookte, placst); there is a 17th century tendency to standardize spelling as -ed, but syncope is indicated in poetry e.g. Dryden’s confess’d (Barber 1997: 175; Nevalainen 2006: 6). Certain Old English strong verbs developed a regular past tense, but both forms remained available in Early Modern English; for example, the past tense of help could be holp or helped, with past participle as either weak holped/helped or strong holp/holpen (with original strong past participle ending -en). Not all strong verbs which developed this past tense variation in Early Modern English (e.g. shake could be shaked or shook) retained the regular form in Modern English (Barber 1997: 174). Some historically weak verbs had strong forms in Early Modern English e.g. snow, snew. Some weak verbs even changed over to the strong class on the basis of analogy e.g. spit and stick (Go¨rlach 1991: 91). Tense marking on strong verbs in Early Modern English often had a different pattern for the form of the preterit and the past participle to both Middle English and Modern English. Different verbs go through different patterns, taking some time to stabilize (Nevalainen 2006: 20). As Lass says “it seems as if each verb has its own history” (1999: 168–170), which can be illustrated by changes in the paradigm for DRINK: late 15th end of 16th to 19th 17th to 19th
drink, drank, drunk drink, drunk, drunk drink, drank, drank
The periphrastic expression of the future with auxiliaries shall and will goes back to Old English. By the early 16th century both auxiliaries had lost much of their modal meaning of obligation and volition and could express pure future.
38 Early Modern English: Morphology Perfect and pluperfect aspect has been expressed through auxiliaries HAVE and BE since Old English (Rissanen 1997: 213); the expression of progressive aspect by means of the BE + present participle construction can also be found in Old English. However, after its growth in Middle English the progressive can only be said to be grammaticalized by 1700, and according to Rissanen (1997: 216), “the set of progressive forms in all tenses, active and passive, is fully developed around the end of the eighteenth century”. He shows how Polonius in Hamlet (II.ii) asks “What do you read my lord?” but in Troilus and Cressida (III.iii) Achilles uses “What are you reading?”. (See further Seoane, Chapter 39.) As always in English, the base form of the verb in Early Modern English serves as the imperative mood. Although in Middle English already there is no distinct plural form of the subjunctive mood, the subjunctive is far more in evidence in Early Modern English than it is in Modern English. This is due in part to the contrast of zero-marked to inflected verb forms in the singular. The subjunctive is typically found in subordinate clauses following a conditional conjunction. In the present, we find the base form of the verb used with the second and third person instead of the inflected forms (-st, -s, -eth). The subjunctive form of BE is invariable be in the present tense (I be, you/thou be, s/he be), and plural were with the singular in the past tense (I were, thou were). This passage (3) from Tyndale illustrates both regular verbs in the subjunctive and BE: (3)
Agre with thyne adversary quicklie / whyles thou arte in the waye with him / lest that adversary deliver the to the iudge / and the iudge delivre the to the minister / and then thou be cast into preson (1526 Tyndale, Bible; Barber 1997: 171)
In modern English traces of the subjunctive remain in phrases such as “long live …”, “if need be” and “if he were”. Through drama especially, it is evident that the subjunctive is not elevated language in Early Modern English, but “comes regularly from the lips of tradesmen, apprentices, artisans, peasants, people with no social pretensions” (Barber 1997: 173). Auxiliaries have been important in the expression of modality since Old English, but the loss of distinctive verb endings almost certainly speeded up the replacement of subjunctive forms by auxiliary periphrasis (Rissanen 1999: 228–230; Nevalainen 2006: 96). For example, we find may used for the optative subjunctive (in heauen may you finde it) and let for hortative subjunctive (let him love his wife even as himself). The preterit subjunctive (were) is replaced by would or should, (4): (4)
if any body should ask me … I should say, I heard so; and it would be very good Evidence, unless someone else were produc’ed (1685 Trial of Titus Oates; Nevalainen 2006: 97)
3 Derivational morphology Both popular and scholarly accounts hold that not only did non-native derivational morphology became productive in the course of the 16th century, but the period showed intensified productivity and creative word-formation with native morphology too. Indeed, it is often remarked that the exploitation of lexical resources in the Renaissance has never been surpassed (Hughes 2000: 162). George Gordon (1928: 262, 269) writes of the “genuine and widespread feeling for word-creation” of the Elizabethans and “the
609
610
V Early Modern English fertility and happy-go-luckiness of Elizabethan English”. Shakespeare’s experiments in word-formation are, for Gordon, the emblem of these Elizabethan tendencies. Scholarly debate has swung between the Victorians who characterized Shakespeare as a Saxonist “lack Latin” who drew mainly on his native vocabulary, and later 20th century critics who claimed that Shakespeare coined thousands of Latinate words. More considered analyses reveal that Shakespeare made extensive use of Latinate prefixes and suffixes, although not always according to the rules of Latin word-formation, for instance in the way that he prefixes the noun moment with in- to form the adjective immoment “unimportant”, or the way that he combines native and non-native elements in hybrids like bi-fold and fore-recited (Garner 1987: 215; Scha¨fer 1973; Scha¨fer 1980). The integration of non-native elements into the English word-formation system began in Middle English, predominantly through the attachment of native suffixes to Fr. bases, for example chasteness (1386). Much less common, and typically later, is the attachment of non-native suffixes to native bases, as in allowment (1579) (Gadde 1910; Nevalainen 1999: 357). Despite their rarity, these hybrid forms are often taken as an indication that lexemes containing the non-native suffix are analyzable for speakers of Early Modern English and that the suffix is thus in some qualitative sense productive (Dalton-Puffer 1996). As most of the new borrowed affixes were in fact limited to Romance and classical bases, it makes sense to speak of a “quantitative shift towards a non-native basis of coining new words in Early Modern English” (Nevalainen 1999: 378). This picture of emerging productivity in non-native affixes in Early Modern English is supported by research following the publication of the Chronological English Dictionary (CED) (Finkenstaedt et al. 1970). With this new tool, Finkenstaedt, Leisi, and Wolff, followed by scholars like Richard Wermser, were able to show how French and Latin loans were the greatest source of new vocabulary between 1600 and 1700 (Finkenstaedt et al. 1973: 118–119; Wermser 1976: 45; Go¨rlach 1991: 166; Nevalainen 1999: 364; Hughes 2000: 152–153). Subsequently it has become clearer that the apparently dramatic peak of Latinate vocabulary observable at the turn of the 16th century is an effect of the OED’s extensive sampling of this period relative to the 18th century (Scha¨fer 1980; Brewer 2006), and in particular the sampling of hard word dictionaries (Osselton 1958; Starnes and Noyes 1946; Barber 1997: 169) (see Lancashire, Chapter 40). Wermser further aimed to show on the basis of the CED how affixation increased in relation to loanwords. Coined words outnumber loans by 58.3% to 37.6% by the 18th century, after two centuries of the two processes being roughly even (Wermser 1976: 40; Nevalainen 1999: 350; Go¨rlach 1991: 138). This proportion is later confirmed by Barber’s 2% sample of the OED (Barber 1997: 221). The relative frequency of nonnative affixes to native affixes in coined words rises from 20% at the beginning of the Early Modern English period to 70% at the end of it (Wermser 1976: 64; Nevalainen 1999: 352). The proportion of Germanic to French and Latin bases in new coinages falls from about 32% at the beginning of the Early Modern period to some 13% at the end (Wermser 1976: 64, 67; Nevalainen 1999: 378). Together these measures confirm the emergence of non-native affixes as independent English morphemes over the Early Modern period. They also seem to contradict claims that the native affixes in Early Modern English are just as, if not more productive, than ever (Barber 1976: 185–188;
38 Early Modern English: Morphology Nevalainen 1999: 391), although it is always less likely that words coined with native affixes would be recorded in a dictionary, especially the Shorter OED, on which the CED is based. We cannot be sure how Wermser was interpreting the etymologies of OED entries – the OED etymologies frequently equivocate, sometimes providing the source of a loan and showing how it could be formed through affixation. For any historical period, it is hard to ascertain whether a given word with a non-native base and a non-native affix is a loan or a coined word, in the “language”, as well as in the mental lexicons of individual speakers. Accounts of Early Modern English word-formation rely on the idea that non-native suffixes become productive over this period, but this is not always based on extensive evidence, and substantial differences in the productivity of processes can be obscured. Thomason and Kaufman (1988) are less persuaded of a new integrated word-formation system emerging in Early Modern English. With the exception of some suffixes like adjective-forming -able (first seen on Middle English loans from French), they consider the derivational phenomena emerging from Latin lexical influence in English post-1450 as “productive for uncultivated speakers to a limited extent only” (Thomason and Kaufman 1988: 308; 1988: 329). Detailed overviews of native and non-native individual prefixes and suffixes can be found in works such as Marchand (1969) and Nevalainen (1999). Like these, the summaries below rely extensively on the OED (Simpson [ed.] 2000–) articles for individual prefixes and suffixes. Here the focus is on affixes emerging in Early Modern English. Sometimes the OED article offers an explicit comment on the stage at which the form is considered to be an independent affix; sometimes this trajectory, where there is one, must be inferred from the dates of coined words. Emphasis is placed on the loanword models for words coined with the new affixes, most commonly on non-native bases. Where non-native affixes do appear on native bases this may be indicative of greater productivity, but not necessarily.
3.1 Prefixation The new negative prefixes, with their general semantics, probably have the greatest impact on the word-formation system of all the new prefixes. Non- is adopted early (late 14th century) through Old French loans which in turn came from Law Latin (nonsense, nonchalant). The prefix first coined words on native and non-native nouns (nontruth; non-activity) but the input range broadened in the 17th century to adjective bases (non-harmonious) including some participles (non-preaching) (Nevalainen 1999: 380) although native adjectival bases (non-bookish, non-English) tend to be 19th century. There are rare examples of non- prefixed to native and non-native verbs (non-act; non-licentiate). In- with its allomorphs il- and im- appears later in the form of Latin (innocens, illiteratus, immensus) as well as Fr. loans (incompetent, inexpressive). From the 16th century we find in- on primarily non-native adjectives (incautious, inarguable, inexpedient; infit). Reversative and privative dis- is also a later addition appearing in Lt. loans such as dispute from disputare even though dis- is not a separable prefix in Latin (Garner 1987: 215). Dis- is described as a “living prefix” after 1600 by the OED, used to form new verbs on existing native and non-native verb bases (disown, disangularise; disrank) and even some noun (discharacter, diseye) and adjective bases (disgood, disrespectful dishonest).
611
612
V Early Modern English All three imported negatives parallel native un-, which appears on all classes of base (unfortunate, unhouse, unnerve), and remains the most common negative prefix in Early Modern English (Nevalainen 1999: 380–382). There is ample evidence of alternation between un- and in- on adjectival bases before the 16th century. The OED indicates that both could appear before the adjectives cautious, ceremonious, certain, communicative, devout, and distinguishable. The practice in the 16th and 17th centuries was to prefer the form with in-, as in inaidable, inarguable, and inavailable, but items with Latinate bases were later revised to in-, with other bases taking un- (unavailing, uncertain, undevout, unexpected). Matthews has described a kind of cyclic process whereby negative prefixes lexicalize with evaluative meanings as in improper, and the alternate prefix remains neutral. Compare unnatural and non-natural, immeasurable and unmeasurable, immoral and amoral (Matthews 1991: 72–74). Words prefixed with in- are probably more inclined to lexicalize in this way given their strong link to Latin lexis. De- and dis- overlap on verb bases as in the oft-cited disthronize, disthrone, dethrone, unthrone, dethronize (Go¨rlach 1991: 80). The prefix de- is only found in the 18th century, although there are some “tentative” 17th century examples like detomb 1607 (Nevalainen 1999: 383). Whilst some suffixes are assimilated relatively early through French, the numerous new prefixes are, by contrast, typically borrowed later from Latin (Burnley 1992: 446–449). They tend on the whole to be restricted to certain technical registers, or at least, to form exclusively technical terms. Typical examples from the set of locatives would be sub- emerging from French loans such as subsequent, subsection, and forming words on all classes of base, as in subtrench, subconsulary, subrenal, and subdecimate; trans-, also from Fr. loans like trespasser and Lt. transferre, forming verbs on verb bases (transplace) and some noun bases (transfashion); and circum-, from Lt. circuminvolvere and circonscrire, appearing on native and non-native verbs (circumbind, circumgyre, circumclose). The intensifying prefix hyper- appears in hyperconformist, hyper-angelical, and hyper-magnetic on analogy of Gk. words like hyperbole, hyperborean. The quantitative prefixes are late 16th or 17th century: multi-, from Lt. multiplex, multifarious and Fr. multiply, multitude, is applied to noun and adjective bases to form multivariety, multilateral; mono-, from Fr. (monarch, monosyllable) and Gk. (monoculus, monoxylon) loans, forms monoptic, monopyrenous; uni-, from Lt. universitas, unicornus, forms univalve, unifoil, unipresence (from which unipresent is then back-derived); bi-, on the analogy of loans bicome, biennium appears principally on non-native adjectives such as bicapsular; tri- appears on noun and adjective bases trigram, tricentrall; Lt. compounds such as semicirculus are imitated to form semi-quaver, semi-riddle, semi-cubit, semi-Atheist. Demi- in fact is somewhat earlier than semi-, appearing in 15th century heraldic loan translations (demigod, demi-angel, demi-lion). A number of the prefixes with productivity restricted by register only show a substantial increase in frequency after the Early Modern period. For example, types like transapical, circumcorneal, postcerebellar, pre-chemical are all 19th century and later. The prefixes pan- (from Gk. pandemic, panoply) and poly- (from Gk. Polygamia) do form words in Early Modern English ( panpharmacon, pantheology, Panglyphic; Polyacoustic) but this is rare, and most examples are 19th century and later. Although pseudo- occurs in borrowed words in Early Modern English ( pseudo-christ from Gk. pseudochristus) it is rarely a “living prefix” in English before 1800 ( pseudo-religious 1672) (Marchand 1969: 188; Nevalainen 1999: 388).
38 Early Modern English: Morphology Some of the new prefixes extend beyond technical terms, and these are often processes that are borrowed earlier. So locative en- became productive in the 15th century and is widely used in the 16th century on native and non-native bases (Nevalainen 1999: 389) to form verbs such as endanger, embody, encamp, ennoble; super-, from Fr. loans superlative, superstition, also takes off in the 15th century and is frequent in Elizabethan times, appearing on nouns (superstructure), adjectives (super-aerial ) and verbs (superinvest); inter-, from Lt. (intercedere, intermedius, interregnum) and Fr. (enterfere, entercourse) loans, leads to formations on native and non-native noun bases (interdispensation, intermatch); native and non-native verb bases (intermention; intertwine) and adjective bases (interconciliary). Temporal prefixes tend to be introduced earlier and found more widely. Re-, from Fr. verbs redress, regard and Lt. reducere (and in contemporary lexicographers’ renderings of Italian words such as ristoppare), becomes “freely prefixed” (OED, Simpson [ed.] 2000–) towards the end of the 16th century, primarily on non-native verbs (re-elect) but also native verbs (regreet). Pre-, from Lt. preambulare, already coins words in late Middle English; these are “numerous” from the 16th century onwards and include pre-petition, pre-excellence on nouns, and on verbs preconceive, preclose, pre-ordinate, pre-sift. Formations after Lt. loans like postponere and Fr. postcommunion, postposer first appear in English in the late 14th century: examples on nouns include post-accession, post-argument, post-pardon and on verbs include postcribrate, post-place. “Attitudinal” prefixes (Nevalainen 1999) tend not to be restricted to technical terms. Counter-, from Fr. counterbalance, countersign, prefixes native and non-native nouns (counterplot, countermotion) and native and non-native verbs (counterhit, counterfix). The Latin version (contraponere) can be found on contra-proposal, contra-civil and contra-distinguish. Anti-, from Gk. (antithesis), appeared exclusively in loan translations such as antipope, antichrist before 1600, but after that was generalized to other noun bases to produce antideity, antiface, antihemisphere, anti-romance and adjectival bases to produce anticreative, antiliturgical.
3.2 Suffixation 3.2.1 Noun suffixes None of the new suffixes forming concrete nouns managed to usurp the ubiquitous native agentive -er suffix (Nevalainen 1999: 392; Go¨rlach 1991: 172). They tend on the whole to be both semantically and formally restricted. -ician is added to arts or sciences in Lt. -ica, Fr. -ique or Eng. -ic, -ics to denote a person skilled in the art or science. Musician and physician are loans but in some cases it is not possible to tell if a word (e.g. magician) is formed in English. Some words like geometrician are formed by analogy on names not even ending in -ic (although there may be an adjective in -ic). -eer is added to English nouns in the early 17th century to form designations of persons ( pamphleteer, auctioneer, pulpiteer) in imitation of earlier Fr. loans like canonnier (> cannoneer) with the Fr. agent suffix -ier (still evident in bombardier). It hardly appears on native bases, and when it does (as in waistcoateer ‘a prostitute’) it is not transparent. Concrete nouns ending in -ant may be Fr. participles borrowed before 1500 (attendant, dependant) later refashioned as Lt. -ent (dependent), or participles borrowed
613
614
V Early Modern English directly from Lt. (stimulant 1728). There are some analogical formations (anaesthesiant 1879) but not many in Early Modern English. Nouns such as curate, senate are English renderings of Lt. nouns curatus, senatus (including medieval Lt. nouns aldermannatus > aldermanate) and this pattern is used to generate words in English on other Lt. nominal stems (syndicate 1624, electorate 1675). Perhaps the most interesting development in this group of noun suffixes is the passive benefactive suffix -ee, for which there is no native equivalent. The first examples are from Anglo-French participles (appellee, refugee) but later words are coined with the suffix in English (referee 1549, vendee 1547). Many subsequent formations in English (laughee 1829) are listed as “nonce-words” but the suffix certainly seems to be alive in Present-day English (Mu¨hleisen 2010). Borrowed abstract noun suffixes are without doubt the most noticeable elements of the new “layer” of derivational morphology. This is due in part to the sheer numbers of complex nouns borrowed, resulting in a wide range of possible noun endings some of which are semantically general. There was already a choice of native abstract-noun forming suffixes in Middle English, particularly for the description of states or qualities as in hethenness, hethenhood, hethenship (Dalton-Puffer 1996: 126). Gerundial -ing is the deverbal noun-forming suffix of choice in Middle English (on native and non-native bases), and the suffix continues to have near inflectional levels of productivity in Early Modern English (Go¨rlach 1991: 172). It is rivalled by the new deverbal suffix -ation, and to a lesser extent -ment (Bauer 2001: 184); other suffixes forming abstract nouns on verbs are more restricted: -ance/-ence became “to a certain extent a living formative” (OED, Simpson [ed.] 2000–) after appearing in Fr. (nuisance, parlance) and Latin or refashioned-as-Latin loans ( providence, prudence) and even coins some nouns on native bases (clearance 1563, hindrance 1436, furtherance 1440); -ance nouns could be refashioned as -ancy if the state/condition meaning was more prominent than the action/process: cf. temperancy 1526 vs. temperance 1340. The suffix -ure (Fr. scripture, Lt. aperture) became “mildly productive” in Early Modern English on verbs ending in -s and -t (Nevalainen 1999: 398) as in exposure 1605; from the 17th century onwards -al from Lt. suffix -alia (via loans like arrival > Anglo-French arrivaille) coins words such as denial 1528. Derivations on native bases (bestowal, betrothal, beheadal ) are all 19th century. Already in the 15th century, -ment is used to coin words denoting the result or product of action or the action itself: chastisement 1340 may be a coined word, and items on Germanic bases like hangment 1440 certainly are. These are modelled on Fr. loans garment, accomplishment and Lt. loans fragment < fragmentum. Later EModE examples include banishment 1507 and enhancement 1577 on Romance bases and amazement 1595 and atonement 1513 on Germanic bases. Some of the latter are also prefixed with em-, en- and be- (enlightenment 1669, bereavement 1731). There are even some formations on adjectives (merriment 1574). The borrowed suffix -ation, however, is considered the most productive deverbal noun-forming suffix after -ing and one of the most productive new suffixes from the Early Modern period. We will examine this suffix more closely to consider what it means to develop productivity in Early Modern English. The productivity of -ation is often attributed to the fact that it is “the only alternative available for verbs ending in -ise, -ate, and -ify” (Nevalainen 1999: 397). Yet some caution is required in treating
38 Early Modern English: Morphology Early Modern words in -ation, even ones on base verbs ending in -ize, -ate and -ify, as confirmation of the emerging productivity of this suffix. English formatives in -ation are considered to “show productivity from the beginning of the 17th century through to the 20th century, but always on Latin or French bases” (Bauer 2001: 181-182) with some well-known exceptions such as starvation. Synchronic morphologists (Kastovsky 1986: 589, 1992: 291) routinely distinguish between -ation words which are recognizable loans such as communion, opinion, protection where -io/-io-n-em has been added in Latin to the stem of a noun (communis), verb (opinari) or participle ( protegere), and the more transparent cases, where Lt. loans such as qualification are formations on the past participial stem of verbs in -are (qualificat- from qualificare). The latter are often treated as English derivations. The general attachment of -ation to non-native bases makes it impossible to tell whether forms which contain the string -ation such as recommendation (a Fr. loan) are the result of borrowing or deverbal derivation in Early Modern English (Nevalainen 1999: 397). Marchand (1969: 259) would like for convenience to treat all items on verb bases in -ate from 1500 as English derivations. So education 1540 would be treated as a derivation even though the OED shows this is a Latin loan. For many of these items the verb is back-derived from the borrowed abstract noun (see verb suffixes below). Sometimes there is not even a back-formed verb to hint at transparency for users as in constellation, duration, ovation (OED -ation article, cited by Marchand 1969: 261). Similarly, Marchand would like to classify -ize + ation words, many of which are Lt. nominalizations, either of Gk. verbs in -ize (baptization) or Lt. verbs in -ize (moralization) or Fr. verbs in -iser (civilization), as English derivations after 1600. We still find loans after 1600 though, such as sacrification 1694. The cut-off of 1600 seems to rather better for -ify + -ation: amplification 1546, modification 1492 and verification 1523 are Latin loans but identification 1644 and beautification 1640 are derivations on verbs in -ify. Interestingly, some early items previously presented by the OED as derivations are now shown as loans for example pontification 1500. More such cases are coming to light in the OED’s latest revisions with the benefit of new resources (Durkin 2002). Finally, there is the question of how we should treat “Latinate coining”, where a noun such as fecundation is in fact formed in English, but on a verb base that exists only in Latin ( fecundare). This is a well-known practice in Early Modern English, yet its extent has not been measured. In sum, the suffix -ation may not be as productive in Early Modern English as is commonly assumed. It might even be argued that this suffix never developed productivity in a quantitative sense. Bauer reflects that recent formations such as lambadazation and electronification must be analogical formations (Bauer 2001: 80–81, 96). The OED in fact indicates that a subset of scientific words including ossification 1671 do not have a pre-existing English verb base. Tellingly, -ization and -(i)fication are listed as complex suffixes alongside -ation. Similar considerations apply to borrowed noun-forming suffix -ity, typically found on non-native adjectival bases in -able/-ible, -ic, -al and -ar and rarely found on native bases (Nevalainen 1999: 398): oddity, the classic exception, is as late as 1713. Unsurprisingly, many of the Early Modern examples turn out to be direct loans from Latin such as implacability 1531, and not a formation from implacable (1552) (Marchand 1969). Lt. nouns in -itas are Englished to -ity often via Fr. -ite. Here too there is the Latinate coining (carneity 1691 is coined in English but the adjectival base carneus does not exist in
615
616
V Early Modern English English) and here too there may be a case for complex suffixes (-ability, -icity) rather than a single -ity suffix. The appearance of native suffix -ness on non-native bases and the consequent appearance of doublets such as sincereness/sincerity; singularness/singularity, fatalness/ fatality (Marchand 1969: 335) is often used to draw attention to affix rivalry in Early Modern English (Nevalainen 1999: 398; Go¨rlach 1991: 137; Romaine 1985; Riddle 1985). Sometimes the increasing productivity of -ity in Early Modern English is presented as claiming territory from -ness (Aronoff and Anshen 1998) but this is based on treating all -ity items as derivations, when in fact many of the rival Early Modern pairs concern an -ity loanword as in absurdity 1529 absurdness 1587 and penetrability 1609 penetrableness 1684. Classical Latin words in -acia ( fallacia > fallacy) or medieval Latin words in -atia (legatia > legacy) are Englished as words ending in -acy. The form is added to Lt. words in -atus (advocatus > advocacy 1413) or English adjectives in -ate (accuracy 1662, privacy 1534) from the 14th century already but is only “generalised” in the 16th century (Nevalainen 1999: 399). The two best known non-native Early Modern English suffixes for forming abstract nouns with a condition /state/ collectivity meaning are -age (from loans such as voyage, umbrage, plumage) and -ery (from loans such as pottery, bravery, machinery). We see -age appear on non-native bases in clientage 1633, orphanage 1538 and non-native bases in leafage 1599, and -ery appears on non-native bases (confectionery 1545) and native bases (brewery 1658). The suffix -ism is striking in that it comes from Gk. loans via Latin (baptism, Atticism, Judaism). From the 16th century it can be found on non-native bases (modernism 1737, magnetism 1616) and native bases (truism 1708). It can simply derive nouns of action ( plagiarism 1621) but its primary uses are semantically narrower: it can denote the conduct of a class of persons ( patriotism 1716), a system of theory or practice (Quakerism 1656), a doctrine or principle (libertinism 1641), or a peculiarity or characteristic (witticism 1677).
3.2.2 Adjectival suffixes As with nouns, numerous adjectives were added to Early Modern English through morphological Anglicization. In many cases an inflectional ending is simply dropped (content < content-us). In others, a set of adjectival loanwords becomes associated with a modified Latinate ending. For example -ary, in Early Modern English appears predominantly in loans such as voluntary and contrary from Fr. voluntaire and Latin contrarius and very infrequently in a word coined in English (complementary 1628). Especially prominent are adjectives in -ate formed from Lt. participles (desolate < desolatus, separate < separatus). Fr. adjectives can be adapted with this ending (affectionate < affectionne`) and so can other Lt. stems (roseate 1589 is from Lt. roseus); thus -ate cannot be considered a productive adjectival suffix. The non-native adjectival suffixes that are productive in Early Modern English and later tend to have gotten off the ground in Middle English. Following Fr. loans such as capable, agreeable, deverbal -able (as noted earlier), which is highly general in meaning, occurs on native (takeable 1449, breakable 1570) as well as non-native bases ( praisable 1350). Whilst new words are coined in this process in Early Modern English, borrowing
38 Early Modern English: Morphology continues. The suffix is attached to nouns from the 16th century: marriageable 1575; but in some cases the base may be the noun or verb e.g. rateable 1503. Deverbal -ive from Fr. (adoptif ) and Lt. (nativus) loans is productively added to Fr. or Lt. verbs, but is formally restricted to those ending in -s or -t (conducive 1646, depressive 1620) as they are essentially analogical formations (Nevalainen 1999: 405); “ative” does become a “living form” as in talkative (1432) but there are few such examples. Denominal -ous (Fr. dangerous; Lt. famosus, obliviosus) is already used to coin words in English from the 14th century (leguminous 1656) although seldom on native bases (timeous 1470), possibly because denominal native adjectival suffixes (muscled, heathery) are widely used in Early Modern English (Nevalainen 1999: 400; Barber 1997: 234). Lt. adjectives in -alem (mortalem) were borrowed early through French with -el (mortel ) later refashioned to -al (mortal ). The number of Lt. adjectives in -alis increased dramatically in medieval and modern Lt. (cordialis) also producing a suffix -al which could be added to any noun (longitudinal 1706; constitutional 1682). The -al ending could also be added to Lt. adjectives with endings such as -eus “to give them a more distinctively adjectival form” (OED, Simpson [ed.] 2000–); e.g. funere-al 1725. In late Lt. -alis nouns (grammaticalis) are formed on adjectives in -ic-us (grammaticus) hence the English grammatical, and so also clerical, medical. Somewhat later Lt. adjectives in -icus are rendered in English with an -ic ending ( poetic < poeticus). Thus we find adjectives with both forms (comic, comical; tragic, tragical ). The historical relationship and semantic differences are explored at length in Kaunisto (2007). Both suffixes occasionally act as independent formatives ( prelatical 1614, operatical 1775) (Nevalainen 1999: 403) but the frequency of this group (Barber 1997 finds -al/-ic/-ical to be the most productive non-native adjectival affix in Early Modern English) is certainly complex. Other adjectival suffixes are semantically narrower and consequently appear on a subset of bases. For instance, -ese (It. Milanese; Fr. Chinois) is added to national proper names only (Japanese); it is extended to other proper names much later (Johnsonese 1843). Similarly, -ian which comes from loans Fr. Barbarien > barbarian and Latin Christianus > Christian is associated with proper names such as Cameronian 1690, despite some Latinate coinings like equestrian 1656 on equestri-s. Whilst -an is added to Lt. adjectives in -arius (agrarius > agrarian) or English adjectives in -ary (disciplinarian), the complex form -ian is mostly associated with ideologies (sublapsarian 1656). There are some jocular formations on native bases in the 18th century (nothingarian 1776). Finally, -ite, which appears in Greek/Latin loans like Israelite, forms person nouns such as Jacobite 1400, Wyclifite 1580.
3.2.3 Verb suffixes Before 1500 the only overt morphological processes available to form verbs were the native prefix be- (bejewel ) and suffix -en (deafen), and the prefix en- (embody) which emerges from Fr. loans in Middle English (endanger). Deadjectival conversions “often compete” with -en suffixations, as in slack and slacken (Nevalainen 1999: 388; 406; 429). Conversion to verb was a much more common process, and so whilst the above verb-forming prefixes were not really in use after 1600, conversion continued and survived into Modern English.
617
618
V Early Modern English Nevalainen (1999: 407) describes -ize as the most productive of the new verb-forming morphological processes of Early Modern English, a situation which continues into Modern English (Plag 1999). This may be partly to do with the fact that -ize appears in relatively fewer Lt. loan words than other borrowed suffixes. Its origins are Greek, from Gk. loans into Latin such as baptize. Because -ize does not appear in so many Lt. loans, most of the -ize words in English such as popularize (1593) are coined, although almost always on non-native bases with some exceptions (womanize 1593). The fashion for -ize verbs attracted controversy in the 16th century, yet they continued to fill up the hardword dictionaries of the 17th century before their demise in the 18th century (Go¨rlach 1991: 176–177). The story of -ify is closer to other Latinate morphology in that most items are renderings of original Lt. verbs in -ficare as in pacify < pacificare; horrify < horrificare). The suffix is also absorbed through Fr. loans (liquefy < liquefier). Coined words such as beautify (1526) are quite rare in Early Modern English. Their addition to native bases is marked as “jocular” or “trivial” (OED, Simpson [ed.] 2000–) in words such as truthify 1647 and speechify 1723. As we saw above, Lt. past participles in -atus, -ata, -atum were a source of English adjectives. Some of these adjectives were treated as verbs (separate 1432). Subsequently English verbs in -ate were formed directly on the Lt. participial stems as in venerate from venerari. In the 16th and 17th centuries some -ate verbs were even coined on Romance nouns (capacitate 1657 from capacity; fertilitate 1634 from fertility), and Latin nominal stems (camphorate 1691 on camphoratus) (Nevalainen 1999: 407). These -ate verbs were stigmatized as “ynkpot termes” in the 16th century. The author of Thomas Wilson’s famous ynkehorne letter from the Arte of Rhetorique (1553) pleads “I obtestate your clemencie, to inuigilate thus muche for me”. Similar items were fabricated by Cockeram in his dictionary of 1623 (Go¨rlach 1991: 176). The exact number of -ate verbs formed through back-formation of -ation nouns, as in locate (1652) from location (1592), is not known, but it is likely to be high throughout the period (Nevalainen 1999: 407; Go¨rlach 1991: 176; Plag 1999). Given the limited productivity of the verb suffixes, it is unsurprising that they are considered to be in complementary distribution (Bauer 2001: 177). Rare “doublets” cited by Plag (1999: 228) (dandify/dandyise; plastify/plasticize) are 19th century. The popular native adverb-forming suffix -ly had already emerged in Middle English. Highly generalized, in Early Modern English it is applied to adjectives (bawdily), including adjectives in -ly (livelily), a practice subsequently discouraged; participles (shortsightedly), numerals (thirdly), and even nouns (agely). However the suffix is less common in adverbs appearing as intensifiers than it is in Modern English (exceeding well ) (Nevalainen 1997: 405).
4 References Allen, Cynthia. 2008. Genitive Case in Early English: Typology and Evidence. Oxford: Oxford University Press. Altenberg, B. 1982. The Genitive v. the of-Construction. A Study of Syntactic Variation in 17th Century English. Lund: Gleerup. Aronoff, Mark and Frank Anshen. 1998. Morphology and the lexicon: Lexicalization and productivity. In Andrew Spencer and Arnold M. Zwicky (eds.), The Handbook of Morphology, 237–248. Oxford: Blackwell.
38 Early Modern English: Morphology Barber, Charles. 1997 [1976]. Early Modern English. 2nd edn. Edinburgh: Edinburgh University Press. Bauer, Laurie. 2001. Morphological Productivity. Cambridge: Cambridge University Press. Brewer, Charlotte. 2006. Eighteenth-century quotation searches in the Oxford English Dictionary. In: R. W. McConchie, Olga Timofeeva, Heli Tissari, and Tanja Sa¨ily (eds.), Selected Proceedings of the 2005 Symposium on New Approaches in English Historical Lexis (HEL-LEX), 41–50. Somerville, MA: Cascadilla Proceedings Project. Brown, Roger and Albert Gilman. 1960. The pronouns of power and solidarity. In: Thomas. A. Sebeok (ed.), Style in Language, 253–276. Cambridge, MA: MIT Press. Brown, Roger and Albert Gilman. 1989. Politeness theory and Shakespeare’s four major tragedies. Language in Society 18(2): 159–212 Burnley, David. 1992. Lexis and semantics. In: Norman Blake (ed.), The Cambridge History of the English Language. Vol. II. 1066–1476, 409–499. Cambridge: Cambridge University Press. Dalton-Puffer, Christiane. 1996. The French Influence on Middle English Morphology: A Corpusbased Study of Derivation. Berlin/New York: Mouton de Gruyter. Durkin, Philip. 2002. Changing documentation in the third edition of the Oxford English Dictionary: Sixteenth-century vocabulary as a test case. In: Teresa Fanego, Bele´n Me´ndez-Naya, and Elena Seoane (eds.), Sounds, Words, Texts and Change: Selected Papers from 11 ICEHL, 65–81. Amsterdam/Philadelphia: John Benjamins. Finkenstaedt, Thomas, Ernst Leisi, and Dieter Wolff. 1970. A Chronological English Dictionary. Heidelberg: Carl Winter. Finkenstaedt, Thomas, Ernst Leisi, and Dieter Wolff. 1973. Ordered Profusion: Studies in Dictionaries. Heidelberg: Carl Winter. Gadde, Fredrik. 1910. On the History and Use of the Suffixes -ery, -age and -ment in English. Ph.D. Dissertation. Lund: Gleerupska University. Garner, Bryan A. 1987. Shakespeare’s Latinate neologisms. In: Vivian Salmon and Edwina Burness (eds.), Reader in the Language of Shakespearean Drama, 207–228. Amsterdam/Philadelphia: John Benjamins. Gordon, George. 1928. Shakespeare’s English. Society for Pure English Tract no. XXIX. Oxford: Clarendon. Go¨rlach, Manfred. 1991. Introduction to Early Modern English. Cambridge: Cambridge University Press. Hughes, Geoffrey. 2000. A History of English Words. Oxford: Blackwell Kastovsky, Dieter. 1986. The problem of productivity in word-formation. Linguistics 24: 585–600. Kastovsky, Dieter. 1992. The formats change – the problems remain: Word-formation theory between 1960 and 1990. In. Martin Pu¨tz (ed.), Thirty Years of Linguistic Evolution: Studies in Honour of Rene´ Driven on the Occasion of His Sixtieth Birthday, 285–310. Amsterdam/Philadelphia: John Benjamins. Kaunisto, Mark. 2007. Variation and Change in the Lexicon: A Corpus-based Analysis of Adjectives in English Ending in -ic and -ical. Amsterdam: Rodopi Kyto¨, Merja. 1996. The best and most excellentest way: The rivalling forms of adjective comparison in late Middle and Early Modern English. In: Jan Svartvik (ed.), Words: Proceedings of an International Symposium, Lund, 25–26 August 1995, 123–144. Stockholm: Kungliga Vitterhets Historie och Antikvitets Akademien. Kyto¨, Merja and Suzanne Romaine. 1997. Competing forms of adjective comparison in modern English: What could be more quicker and easier and more effective? In: Terttu Nevalainen and Leena Kahlas-Tarkka (eds.), To Explain the Present. Studies in the Changing English Language in Honour of Matti Rissanen, 329–352. Helsinki: Socie´te´ Ne´ophilologique. Lass, Roger. 1999. Phonology and morphology. In: Lass (ed.), 23–155. Lass, Roger (ed.). 1999. The Cambridge History of the English Language. Vol. III. 1476–1776. Cambridge: Cambridge University Press.
619
620
V Early Modern English Marchand, Hans. 1969. The Categories and Types of Modern English Word Formation. 2nd edn. Wiesbaden: Otto Harrassowitz. Matthews, Peter H. 1991 [1974]. Morphology. 2nd edn. Cambridge: Cambridge University Press. Mu¨hleisen, Susanne. 2010. Heterogeneity in Word-Formation Patterns: A Corpus-based Analysis of Suffixation with -ee and its Productivity in English. Amsterdam/Philadelphia: John Benjamins. Nevalainen, Terttu. 1999. Early Modern English lexis and semantics. In: Lass (ed.), 332–458. Nevalainen, Terttu. 2006. An Introduction to Early Modern English. Edinburgh: Edinburgh University Press. Nevalainen, Terttu and Helena Raumolin-Brunberg. 2003. Historical Sociolinguistics: Language Change in Tudor and Stuart England. London: Longman. Osselton, Noel E. 1958. Branded Words in English Dictionaries before Johnson. Groningen: Wolters. Plag, Ingo. 1999. Morphological Productivity: Structural Constraints in English Derivation. Berlin/ New York: Mouton de Gruyter. Riddle, Elizabeth. 1985. A historical perspective on the productivity of the suffixes -ness and -ity. In: Jacek Fisiak (ed.), Historical Semantics: Historical Word-formation, 435–461. Berlin/New York: Mouton de Gruyter. Rissanen, Matti. 1999. Syntax. In: Lass (ed.), 187–331. Romaine, Suzanne. 1985. Variability in word formation patterns and productivity in the history of English. In: Jacek Fisiak (ed.), Papers From the 6th International Conference on Historical Linguistics, 451–65. Amsterdam/Philadelphia: John Benjamins. Scha¨fer, Jurgen. 1973. Shakespeares Stil: Germanisches und ronianisches Vokabular. Frankfurt: Athenaum. Scha¨fer, Jurgen. 1980. Documentation in the OED: Shakespeare and Nash as Test Cases. Oxford: Oxford University Press. Simpson, John (ed.). 2000–. The Oxford English Dictionary. 3rd edn. online. Oxford University Press. www.oed.com/ Starnes, Dewitt T. and G. E. Noyes. 1946. The English Dictionary from Cawdrey to Johnson, 1604– 1755. Chapel Hill: University of North Carolina Press. Stein, Dieter. 1988. On the mechanisms of morphological change. In Michael Hammond and Michael Noonan (eds.), Theoretical Morphology: Approaches in Modern Linguistics, 235–249. London: Academic Press. Taylor, E. W. 1976. Shakespeare’s use of -eth and -es endings of verbs in the First Folio. CLA Journal 19(4), 437–457. [Reprinted in Vivian Salmon and Edwina Burness (eds.), Reader in the Language of Shakespearean Drama. Amsterdam/Philadephia: John Benjamins, 1987.] Thomason, Sarah G. and Terrence Kaufman. 1988. Language Contact, Creolization and Genetic Linguistics. Berkeley: University of California Press. Wales, Katie. 1983. Thou and you in Early Modern English: Brown and Gilman re-appraised. Studia Linguistica 37(2): 107–25. Wermser, Richard. 1976. Statistiche Studien zur Entwicklung des englischen Wortschatzes. Bern: Francke.
Claire Cowie, Edinburgh (UK)
39 Early Modern English: Syntax
621
39 Early Modern English: Syntax 1 2 3 4 5 6 7 8 9
Introduction Genitive The loss of impersonal constructions Changes in verbal periphrases Changes in the complementation system Negation The regulation of word order Conclusions References
Abstract Early Modern English is an important period of transition between a still largely synthetic language heavy with variants and a fairly analytic, standardized one. The study of EModE syntax benefits greatly from the abundant and heterogeneous linguistic data available in this period, which allows for the inclusion of sociolinguistic and stylistic factors in the analysis of the changes. In this chapter I concentrate on the description of those which are considered quintessential to the period. These are changes in relation to the long-term transformation of English from a synthetic to an analytic language, such as major developments in the verbal system, as well as changes indicative of the emergence of the written standard, with a reduction of variation and the establishment of rules of usage in areas such as negation and word order. Finally, I also focus on those changes which first emerged in this period and which are still developing today, such as the restructuring of the complement system.
1 Introduction The works of William Shakespeare, Ben Jonson, John Dryden, and many other writers provide evidence of considerable syntactic changes during the EModE period. For example, in Hamlet, Shakespeare has Bernardo ask Horatio “What think you on’t?” (I.i), while Lord Polonius asks King Claudius “What do you think of me?” (II.ii). A wealth of such variants, most of them inherited from Middle English, characterize the beginning of the period, whereas by the turn of the 18th century variation has been greatly reduced, and the language that then emerges is quite similar to the standard we have today. The 16th and 17th centuries witness the establishment of the written standard, and therefore it comes as no surprise that during this period rules of usage and functional differentiations become established and settled in many areas of English syntax. Changes conforming to this general process of regularization include the differentiation between the preterit and the perfect (Section 4.1) and the fixation of new word order rules (Section 6). Other changes, however, are merely incipient in our period, and would only develop fully in the 19th or 20th century; these include the encroachment of the gerund on the territory of the to-infinitive within the complement system (Section 5). Most of the changes observed in Early Modern English are in direct relation Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 621–637
622
V Early Modern English to the transformation of English from a synthetic to an analytic language which started in Old English: the loss of case and verb inflection implied a greater development of the verbal system, for example in the functional and formal expansion of the progressive periphrasis (Section 4.2), the introduction of an obligatory subject, with the consequent loss of impersonal constructions (Section 3), and the fixation of SVO word order (Section 7). The sources of data for the study of syntactic change in Early Modern English are abundant and offer a wide range of styles and registers, especially in comparison with Old and Middle English. Though access to truly oral data, the initial locus of change, is not possible, corpora such as the Helsinki Corpus (henceforth HC) (Rissanen et al. 1991), the Corpus of English Dialogues (1560–1760), the Corpus of Early English Correspondence Sampler, the Parsed Corpus of Early English Correspondence, the Lampeter Corpus of Early Modern English Tracts, the Penn-Helsinki Parsed Corpus of Early Modern English, and the first part of ICAMET, that is, the Innsbruck Letter Corpus (1386 to 1688), allow the researcher to form hypotheses about the sociolinguistic factors conditioning the development of particular changes. This chapter presents a selection of the major syntactic changes in the EModE period, excluding two which have been dealt with in their own chapters, the development of periphrastic do (see Warner, Chapter 47) and changes in relative clauses (see Johansson, Chapter 49). Examples provided are taken from studies of the particular phenomena under discussion as well as from the HC (Rissanen et al. 1991). For the many other interesting syntactic changes that have necessarily been left out of this chapter, such as the restrictions in the use of adjectives as head of NPs, the decline of the subjunctive, the development of modal auxiliaries, the loss of reflexive verbs, the loss of the use of “pleonastic“ that after adverbial conjunctions, the variation and changes in the realm of causative and conditional links, the reader is referred to the exhaustive Cambridge History of the English Language chapter by Rissanen (1999: 187–331).
2 Genitive One of the most surprising changes in the syntax of NPs in English takes place in our period, namely the revival of the s-genitive. In Old English the productive synthetic genitive or s-genitive (as in PDE Mary’s car) was used to express many different types of relations between the head and modifier, while the analytic of-construction (the wheels of the car) was only marginal at best. In Middle English, in compliance with the long-term shift of English from a synthetic to an analytic language, the of-construction gains ground and takes over most of the functions of s-genitives. Contrary to what might be expected, however, after a long period of replacement by the of-genitive, the s-genitive in Early Modern English increases its frequency again. This recent finding by Rosenbach and Vezzosi (2000) and Rosenbach et al. (2000), further discussed in Rosenbach (2002: Chapter 7), comes to contradict the widely established view that the replacement of the s-genitive by the of-genitive which started in Middle English continued in Early Modern English forming a typical S-curve process (Altenberg 1982: 302; Rissanen 1999: 201). In this period there is also evidence of a change of status for the ’s, from an inflection to a clitic-like element. Similarly, the s-genitive in this period develops a new function,
39 Early Modern English: Syntax that of a definite determiner. Both developments, which are discussed and justified in Rosenbach (2002: 201–230), paved the way towards a more frequent use of the s-genitive. The variation between the two constructions in our period is determined by the following (hierarchically-ordered) factors: animacy, topicality, and type of relations. This way, the s-genitive is restricted to human and animate possessors exclusively, and is likewise favored for highly topical possessors (i.e. referentially given and definite) and for the expression of typically possessive relations between the NPs concerned. Other interesting developments in the realm of the genitive is the emergence of the group genitive, as in (1), and the spread of the so-called double genitive and absolute genitive, as in (2) and (3), both of which had emerged in Middle English (cf. Rissanen 1999: 202–204): (1)
They met two of the king of Spaines armadas or Gallions (1600 Chamberlain 94; Rissanen 1999: 202)
(2)
He keeps her the prettiest pacing Nag with the finest Side-saddle of any Womans in the Ward (1672 Shadwell 128; Rissanen 1999: 204)
(3)
Where did he lodge then? […] At Mr. Jyfford’s, or Mr. Harwell’s (1680 The Trial of Titus Oates, P IV, 82.C1; HC, Rissanen 1999: 203)
3 The loss of impersonal constructions Much has been written about impersonal verbs and impersonal constructions in the history of English (cf. especially Elmer 1981; Fischer and van der Leek 1983, 1987; Denison 1990, 1993; Ogura 1990; Allen 1995; Fischer et al. 2000). For obvious reasons of space, I will not discuss the different classifications and theories put forward in the literature (let alone the terminological entanglement) regarding impersonal verbs, but will concentrate on the state of affairs in Early Modern English, which sees the decline and disappearance of the last impersonal verb constructions. While in Old and Middle English constructions without a syntactic (nominative) subject were common, their use decays rapidly in late Middle English, and is replaced by the use of constructions with dummy subject it or with a non-experiencer subject. In Early Modern English both types of structures can be found: impersonal verbs without an overt subject, as in (4) below, or with dummy it or a non-experiencer NP as subject, as in (5) and (6). (4)
And therfore me semeth beste to holde my peace, least I shoulde do as the knyght of the toure dyd (1534 Anthony Fitzherbert, The Book of Husbandry 97; HC)
(5)
whiche your worshipfull benignitee, could sone impetrate for me, if it would like you to extend your scedules, and collaude me in them, (1553 Sir Thomas Wilson, The Arte of Rhetorique; Go¨rlach 1991: 221)
(6)
this lodging likes me better (1599 Shakespeare, King Henry V IV.i; Brinton and Arnovick 2011: 369)
623
624
V Early Modern English Subjectless structures such as (4) decline and eventually disappear in the 16th century, and, as already mentioned, are replaced either by constructions of the type illustrated in (5) and (6) (me likes this > it/this likes me) or by a personal construction (I like). Around the 17th century, impersonal constructions with it as subject also disappear, though Present-day English retains a few, such as it seems to me that or it happened that (Brinton and Arnovick 2011: 369). A number of subjectless phrases, among them methinks, methought, and meseems, deserve special mention. These survive longer than other impersonal verb constructions probably because they acquire adverbial status in Early Modern English and have already become stereotyped by the 16th century. With regard to methinks, the most frequent of all, it loses its original compositional meaning denoting a process of cognition, and develops interpersonal meaning, marking evidentiality; in fact, no pronouns other than me are attested in this combination in Early Modern English (*him thinks; cf. Rissanen 1999: 250–251). The occurrence of formations such as my thought(s) or methoughts (cf. (7) below) show that methinks is no longer perceived as the combination objective pronoun me and impersonal verb thinks. (7)
Methoughts that I had broken from the Tower And was embark’d to cross to Burgundy; (1597 Shakespeare, Richard III IV.i)
For some scholars methinks undergoes a process of grammaticalization in our period (cf. Brinton 1996; Palander-Collin 1999), while for others it is a combination of grammaticalization and lexicalization (Wischer 2000; cf. also Lo´pez-Couso 1996). The traditional explanation for the loss of impersonal verbs, as found, for example, in Jespersen (1909–49: III, 11.2), involves the reanalysis of the (normally) preverbal oblique pronoun as subject (Me likes it > I like it), triggered by the loss of inflections, which would render many structures ambiguous, and the fixation of SVO order. However, this traditional account has faced sound criticism by those who argue that reanalysis would entail variation between speakers, whereas what evidence shows is variation across lexical verbs (Allen 1995: 450–451). Their criticism is also based on the fact that empirical data reveal that the proportion of alleged ambiguous constructions resulting from loss of inflections is notably low (cf. Allen’s 1986 paper on like; see also von Seefranz-Montag 1984). Moreover, the reanalysis theory would not account for the emergence of new impersonal verbs in Middle English (such as remember) and the development of impersonal uses of already existing verbs (such as behove, see Allen 1997; for a full discussion of the criticism to the traditional view, cf. Loureiro-Porto 2010). This is undoubtedly a fascinating area of research in need of further insights.
4 Changes in verbal periphrases 4.1 The perfect be/have + past participle The perfect periphrasis in Early Modern English differs from Present-day English in two broad respects: firstly, with regard to the function for which it is used, and secondly, regarding the auxiliary it takes, which can be either have or be. Illustrative of the first difference are the following examples:
39 Early Modern English: Syntax (8)
I have delivered it an hour since (c.1601 Shakespeare, All’s Well that Ends Well IV. ii; Elsness 1997: 250)
(9)
You spoke not with her since? (c.1608 Shakespeare, King Lear IV.iii; Barber 1997: 190)
In (8) the perfect is used where today we would find a preterit (I delivered it an hour ago), while the reverse situation is found in (9) (You have not spoken with her since?). These two examples show how the perfect, which originated in Old English and became part of the tense and aspect system in Middle English, competed with the preterit for the expression of past time in Early Modern English. In other words, the clear-cut functional differentiation between the two had not yet been reached; in fact, the PDE rules for their use became established only in the early 18th century (Fride´n 1948: 27–37; Go¨rlach 1991: 111; Rissanen 1999: 224–227; Fischer and van der Wurff 2006: 140–141). Regarding the auxiliaries that could mark the perfect in Early Modern English, we find variation between have and be, the two auxiliaries which, since OE times, could be used for that purpose (cf. Brinton 1988: 99–102). The variation between them at this point can be described as follows. Have is normally selected with transitive verbs, and is also predominant with intransitive verbs if these are non-mutative. Mutative intransitives (denoting change of state or place) alternate have and be (cf. 10 and 11), but be is much more frequent. According to Kyto¨ (1997), have occurs in 95% of the perfect periphrases with non-mutative intransitives, while it occurs in only 30% of the periphrases with mutative intransitives. (10) and then if she find us on the Bed, she will verily conceive that we have gone astray, and Erred from the Light. (1633–1703 Pepys, Penny Merriments 148; HC) (11) Wough, she is gone for euer, I shall hir no more see. (1533 Udall, Roister Doister l. 1077; HC) The factors influencing the choice between the two auxiliaries in intransitives include the author’s idiolect as well as a number of linguistic factors. The most important of these is the intended meaning: if the focus is on the action expressed by the verb, have is preferred, whereas be tends to be used to highlight the state resulting from the action (cf. 10 and 11). Due to the association between have and action contexts, have also tends to be singled out in (i) irrealis or hypothetical contexts (e.g. conditional clauses, counterfactuals), where a resultative state is not normally reached and the emphasis is therefore on the (hypothetical) action; and (ii) in iterative and durative contexts (e.g. with an adverbial expressing duration) (cf. Fride´n 1948: 44–57; Denison 1993: 355–356; Rissanen 1999: 213–214; Fischer and van der Wurff 2006: 141–142). Through the period, have progressively gains ground at the expense of be. Some scholars, such as McWhorter (2002), attribute this process to Scandinavian influence. However, the reason most commonly adduced for the disappearance of be as a perfect auxiliary in English is the ambiguity derived from the fact that be-perfects are identical
625
626
V Early Modern English to passive periphrases, which in Early Modern English were also expressed with the combination be + past participle. That both is and has could appear as ’s added to the confusion between both periphrases (Rissanen 1999: 215; Fischer and van der Wurff 2006: 142). Despite the steady retreat of be, it is still predominant in mutative intransitives at the end of our period and will only be superseded by have in the early 19th century (Kyto¨ 1997: 19), except for some relics where be is still found today (as in Dinner is served, cf. Brinton and Arnovick 2011: 373).
4.2 The progressive be + -ing The progressive (She is reading a book) was established in late Middle English (cf. Fischer 1992: 250–256), but it is in Early Modern English when its frequency increases and its use is expanded to cover nearly all the functions it has in Present-day English, as has been shown conclusively in a number of studies (Nehls 1988; Denison 1993: Chapter 13; Elsness 1994; Nu´n˜ez-Pertejo 2004). The origin of the progressive is highly disputed. There is no agreement as to whether it derives from one of the following OE constructions or from a combination of both: (i) be and the present participle ending in -ende, and (ii) the construction made up of be, the preposition on, and a gerund in -ing/-ung, which conveys a meaning similar to that of the progressive (cf. Fischer 1992: 250–256; Traugott 1992: 187–190; Denison 1993: 400–408). Our period witnesses the last remnants of the gerundial construction and a crucial development of the progressive be + -ing. The grammaticalization of the progressive, however, is not complete in Early Modern English but takes place later, during the Late Modern English period (cf. Nehls 1988: 183; Denison 1993; Smitterberg 2005; Fischer and van der Wurff 2006: 136). In this section I will first illustrate the use of the gerundial construction in Early Modern English, and then concentrate on the development of the progressive in Early Modern English and the differences it exhibits with Present-day English, as regards both its meaning and its paradigm. As already mentioned, the OE construction with be + on + verbal noun in -ing still lingers on in Early Modern English, specially in colloquial styles. Nehls (1988: 184) reports, for instance, that while it does not occur in the elevated style of John Evelyn’s Diary, it is relatively frequent in the diary of Samuel Pepys, this latter being more colloquial in nature. Similarly, in the EModE section of the Helsinki Corpus (HC) (Rissanen et al. 1991) this construction features exclusively in fiction, private letters and comedies (Rissanen 1999: 217; Nu´n˜ez-Pertejo 2004: 153–154). Another characteristic of the gerundial construction is that it generally contains either an intransitive verb (very commonly one of motion, cf. 12) or a verb with passive meaning (as in 13). In both cases the verb is not followed by an object, and this tendency to appear without an overt object is interpreted by Elsness (1994: 22) as caused by the nominal origin of the -ing form involved. As illustrated by Nehls (1988: 184, cf. example 14), however, these gerundial constructions have been attested with objects as well. (12) Whither were you a-going? To the Cardinal’s; (c.1613 Shakespeare, King Henry VIII I.iii.50; Nehls 1988: 184)
39 Early Modern English: Syntax (13) Yr gowne and things are a making, but will not be done against whittsunday (1620– 44 Thomas Knyvett, The Knyvett Letters 57; HC, Elsness 1994: 16) (14) I kill’d the slave that was a-hanging thee (c.1608 Shakespeare, King Lear V.iii.274; Nehls 1988: 184) It is not only a, the weakened form of on, that is possible in our period; other prepositions, such as in or upon, are also recorded, though at lower frequencies (cf. Elsness 1994: 13; Rissanen 1999: 217; Nu´n˜ez-Pertejo 2004: 152–156). The development of the progressive in this period is remarkable on several counts. For one thing, its frequency more than tripled (from 16.8 occurrences per 100,000 words in 1500–1570, to 55.5 in 1640–1710), being considerably more frequent in informal than in formal registers: 53.8 tokens per 100,000 words as against 25.0 (data from the Helsinki Corpus in Nu´n˜ez-Pertejo 2004: 161, 173; see also Elsness 1994: 11). As we know, this trend toward a frequent use and a preference for colloquial registers would continue to characterize progressives in Late Modern and Present-day English (Denison 1998). Closely related to this sharp rise in frequency is the fact that in Early Modern English the progressive begins to function as an aspectual marker, mainly indicating limited duration, as illustrated in (15). (15) That done, I will be walking on the works; (1603 Shakespeare, Othello III.ii.3; Nehls 1988: 182) Together with the unequivocal expression of aspect, most of the functions that characterize the modern progressive were present in Early Modern English, such as the expression of futurity (I am leaving tomorrow), the use of the progressive with always (He is always doing that) and the so-called interpretative progressives (Smitterberg 2005: Chapter 7; for a detailed discussion of the semantics of the progressive cf. Brinton 1988: 7–10, 38–45; for Early Modern English in particular cf. Nu´n˜ez-Pertejo 2004: 177–191). Despite its highly significant development, the progressive in Early Modern English still differed from Present-day English in a few important respects. Firstly, in contrast to Present-day English, the use of the progressive for the expression of limited duration was only optional, which shows that the grammaticalization of the periphrasis was not yet complete (Nehls 1988: 181–182; Fischer and van der Wurff 2006: 136). Sentence (16), for example, illustrates the use of a simple present to express limited duration, which would not be possible today. Another example of the unsettled use of the progressive in Early Modern English is (17), where a progressive is employed instead of a simple present form to express unlimited duration, possibly for stylistic purposes (cf. Ryde´n 1997: 422). Finally, (18) instantiates the use of the progressive with a stative verb, which would be highly unexpected in Present-day English. (16) No! what does she here then? Say, if it be not a woman’s lodging, what makes she here? (1675 Wycherley, The Country Wife 135; Nehls 1988: 182) (17) And first we cam to Torrens Cedron, which in somer tyme ys Drye, And in winter, and specially in lente, it ys mervelows flowing with rage of water that comyth with
627
628
V Early Modern English Grett violence thorow the vale of Josophat. (1517 Torkington, Ye Oldest Diarie of Englysshe Travel 27; HC, Nu´n˜ez-Pertejo 2004: 165) (18) I know you expect I should tel you what is become of the money I brought along with me: and I will gladly satisfy you in any thing. Some of it is yet remaining in my hands, for uses: (1643–1737 Strype, Letters 181; HC, Elsness 1994: 20) Another relevant difference between Early Modern English and Present-day English is the fact that the paradigm of the progressive was still incomplete, since the progressive passive (She was being arrested) is not recorded until the end of the 18th century. Instead, the EModE speaker would employ an active progressive with passive meaning, as in (19). (19) Also, they told us for certain that the King’s statue is making by the Mercers Company (who are bound to do it) to set up in the Exchange. (1660 Pepys, Diary I 113.26; Denison 1993: 390–391) These constructions, commonly called “passival” (cf. Visser 1963–73: Sections 1872– 1881), have an inanimate subject which would correspond to the subject in a true passive and to the object in a true active. Passival constructions survived well into the 19th century (Denison 1998: 151), and there are some remnants in Present-day English (as in This movie is shooting in Vancouver, cf. Brinton and Arnovick 2011: 374). Progressive passive import could also be conveyed by gerundial constructions, as illustrated in (13) above (your things are a making ‘are being made’). The availability of these alternative forms of expressing progressive passive meaning delayed the development of the progressive passive periphrasis, a development which was probably resisted also because it involved the progressive of the verb be, not available until Late Modern English (cf. Denison 1998: 146–147), and the occurrence of two consecutive auxiliaries be, which did not take place in any other context (Brinton and Arnovick 2011: 374). The next period would see the continuation of the trend that started in Early Modern English towards a multifunctional and more frequent use of the progressive, especially in informal registers. Also, Late Modern English would host other changes in the progressive leading to its final grammaticalization, such as a more frequent use of the progressive with inanimate subjects, its expansion to non-dependent clauses, the obligatorification of its use as an aspectual marker, and the emergence of the progressive passive.
5 Changes in the complementation system Early Modern English marks the beginning of a series of changes that would lead to the “massive restructuring of the complement system” (Fanego 2007: 162) which takes place in modern times. Such restructuring involved radical changes in the inventory, frequency and distribution of the different types of sentential complements available in the language. For reasons of space, I will concentrate on the most relevant of these changes: (i) first and foremost, the emergence and spread of gerundive clauses, which would eventually encroach upon the territory of to-infinitives in both subject and object positions; and (ii) the spread of the recently established for … to construction to new environments.
39 Early Modern English: Syntax While all major types of sentential complements date back to Old English, gerundive clauses with and without a subject (cf. 20 and 21 below) emerge in medieval times. (20) John resents my / me working in a bar (21) Accepting the job was a good idea The gerunds in these clauses have a nominal origin; the -ing suffix comes from the OE derivational suffix -ing / -ung, which was used to create deverbal nouns of action, such as OE sceawung ‘observation’ < sceawian ‘observe’ (Kastovsky 1992: 388). In Middle English -ing nouns often occurred in phrases which were ambiguous between a nominal and a verbal reading, since they lacked determiners and involved constituents that could occur in NP or VP structure, such as (22) below. As Fanego (2004: 18–26) demonstrates, this is the locus for the reanalysis of the ME noun in -ing as a verb, which took place in late Middle English. (22) Vnder þe Monument ʒeo stod wiþoute wepyng sore (c.1280 Southern Passion 1874; Tajima 1985: 101) ‘she stood close by the sepulcher without weeping bitterly / without bitter weeping’ Though the first signs of the verbalization of the gerund can be traced back to Middle English, its use with fully verbal characteristics was not systematic until the end of the EModE period (cf. Tajima 1985; Donner 1986; Fanego 1996b). Consequently, in Early Modern English gerunds can be fully nominal, as in (23) below, fully verbal, as in (24), or appear in a number of hybrid constructions, which combine nominal and verbal characteristics, as in (25), where the gerund has nominal premodification and a direct object. (23) the maine point belonging therunto is the Hus-wiues cleanlinesse in the sweet and neate keeping of the Diary House. (1615 Markham, Countrey Contentments 109; HC, Fanego 1996b: 97) (24) the whole nation now exceedingly alarm’d by the French fleete braving our Coast even to the very Thames mouth: (1689–90 Evelyn, Diary 927; HC, Fanego 1996b: 97) (25) to adore that great mystery of Divine Love (which the Angels, better and nobler Creatures than we are, desire to pry into) God’s sending his onely Son into the world to save sinners (1671 Tillotson, Scoffing at Religion 429; HC, Fanego 1996b: 97) Closely related to the emergence and spread of verbal gerunds is the fact that gerunds experience a considerable rise in their frequency of use, especially those occurring in constructions which mirror VP structure, that is, gerunds containing post-head dependents. On the contrary, gerunds with only pre-head modifiers decrease in frequency, which is clearly linked to the increasing verbalization of gerunds in our period (Fanego 1996b: 115; 2004: 11–18).
629
630
V Early Modern English The verbalization of gerunds took place first in prepositional environments, illustrated in (26) below. Other environments, such as subjects or objects, were often blocked by the productive to-infinitive, which nevertheless could not occur after prepositions other than to. From this prepositional environment, however, verbal gerunds soon spread to other syntactic positions, so that by the end of the EModE period the verbalization of gerunds had nearly reached completion in all syntactic contexts (cf. Fanego 1996b: 125; 2007: 169–170). (26) This Vlixes […] callis hym the cavse of cacchyng this toun This Ulysses calls him the cause of capturing this town (?a1400 ‘Gest Hystoriale’ of the Destruction of Troy 12204; Tajima 1985: 76) Thus, the 16th century sees the first clear cases of verbal gerunds as sentential complements, in particular as objects of subject-control verbs, as in (27). So-called subjectcontrol verbs (such as refrain below) govern a complement clause which has an unexpressed subject (PRO) whose antecedent or controller is the matrix subject (he). (27) They come so to purpose, that hee can not refraine telling them. (1561 T. Hoby tr. Castiglione’s Courtyer (1577) D iv; OED, s.v. refrain v., def. 5b [Fanego 1996a: 38]) After our period, verbal gerunds extended their use to new environments, where only to-infinitives had been possible before. In other words, Early Modern English hosts the beginning of the process whereby gerundives progressively gain ground at the expense of to-infinitives, a process which seems to have reached its highest peak in the 19th century but still goes on today (Denison 1998: 256). Fanego (2007) shows that this process can be characterized as a drift (in Sapir’s 1921 terms), that is, as “a long-term succession of changes […] towards a greater specialization of the infinitive, which has largely come to be used in complementary distribution with the gerundive” (Fanego 2007: 162–163). Early Modern English also witnesses significant changes in another type of complement, namely the for NP to-infinitive type, as in (28). (28) Water of mynte […] were good for my cosyn to drynke for to make hym to browke (1473 Marg Paston Lett. III; Cuyckens and De Smet 2007: 90) The origin of this construction lies in the reanalysis of the ME sequence [for NP] [to-infinitive], where for is a preposition introducing a benefactive NP followed by a toinfinitive. This sequence is reanalyzed into a complementizer for introducing the nonfinite complement [NP to-infinitive], in which the NP functions as subject of the following to-infinitive (cf. Fischer 1992: 330–333; Cuyckens and De Smet 2007: 92–94). The first unambiguous examples date back to late Middle English, cf. (28) above. However, it is in Early Modern English when this construction becomes frequent (around 20 occurrences per 100,000 words) and acquires considerable functional development. Originally, the most frequent function is that of extraposed subject (28), which reached its most marked increase in Early Modern English, but in this period the construction also extended to function as subject (29), adjunct (30) and a myriad of other functions (for a discussion of the motivations behind the functional spread, cf. Cuyckens and De Smet 2007).
39 Early Modern English: Syntax (29) and for the quantity of milke, for a Cow to giue two gallons at a meale, is rare, and extraordinarie; (1615 Markham, Countrey Contentments, 613; HC, Cuyckens and De Smet 2007: 99) (30) Item that the Wever whiche shall have the wevyng of eny wollen yerne to be webbed into cloth shall weve werk and put into the webbe for Cloth to be made therof (1511–12 The Statutes of the Realm III; HC, Cuyckens and De Smet 2007: 101)
6 Negation This section concentrates on the development of negation in Early Modern English, which is closely related to the development of the dummy auxiliary do (cf. Warner, Chapter 47). The term that best captures the essence of negation in Early Modern English is variation, since several negation patterns, old and new, converge and coexist in this period. Starting with the oldest patterns, the OE and ME negative form ne is still recorded in some examples, sometimes on its own but normally together with not (from OE nawiht). (31) Ne they be not in commune […] nor one man hath nat al vertues (1531 Elyot, The Boke named The Governour; Barber 1997: 198) Not starts to be used in Middle English in combination with ne in order to reinforce the negation and becomes very frequent in Early Modern English. The result is multiple negation, as shown in (31), which can often accumulate three or four negative words in a clause. Multiple negation is very much reduced in frequency in 17th century written English, but survives to face the prescriptive fury of early grammarians in the 18th century and, as we know, it is still amply recorded in Present-day non-standard English (Denison 1998: 243; Rissanen 1999: 272; Mazzon 2004: Chapter 5). Along with multiple negation, cases of negation using only not can also be found in Early Modern English. The original placement of not is postverbal (cf. 32), given its reinforcing function, but after the disappearance of ne the postverbal position was probably felt to be “unnatural” for not, since (i) negatives were traditionally associated with preverbal position (where ne used to appear), (ii) not in postverbal position separated the transitive verb from its object in a period when fixation of SVO word order was taking place, and (iii) there was a general tendency in our period to move light adverbs to preverbal position (Ellega˚rd 1953: 194; Rissanen 1999: 267–268). For this reason, variation between the old postverbal and a new preverbal position for not, as in (33), is common: (32) Wednesday, the vj Day of Januarii, the wynde Rose a yens vs, with grett tempest, thonnderyng and lyghtnyng all Day and all nyght, So owtrageowsly, that we know not wher wee war. (1517 Torkington, Ye Oldest Diarie of Englysshe Travell 60; HC) (33) I not doubt He came aliue to Land (c.1610 Shakespeare, The Tempest I.i; Rissanen 1999: 271) The most commonly occurring pattern, however, was auxiliary -not -V, as in (34) below, probably because it preserved the natural preverbal position for the negative particle
631
632
V Early Modern English and, in addition, did not separate the subject from the verbal element carrying tense and number. This is the pattern where the relationship between the development of do in negatives and the preference of not for preverbal position seems to be one of mutual dependency and reinforcement (Denison 1993: 467; Rissanen 1999: 271; Fischer and van der Wurff 2006: 157–158). (34) Mr Edmondes. Vntill this very day wee haue not heard one worde of yow since your departure, wch kept vs in douptfulnes of your safetie (1570–1640 Robert Cecil, Letters; HC) Other negative elements expressing negation alone or in combination with each other (and with not) are nothing, none, or never, cf. (35). (35) But if ye will sel it, send word to your son what ye will doe, for I know nothing els wherwith to help you with (1500–70 Isabel Plumpton, Letters to Husband; HC) The occurrence of examples like this, where never or nothing occur as sole negators, have led scholars to believe that these forms were on their way to becoming bleached and grammaticalized, just as not resulted from the grammaticalization of the OE emphatic negative particle nawiht ‘nothing’. While these forms did undergo bleaching in other varieties (e.g. never / ne’er in some contemporary non-standard British dialects, cf. Cheshire 1998: 129–130), their grammaticalization would have been blocked in Early Modern English by the fixing of the standard (Blake 1983: 110; Mazzon 2004: 61). Another difference between Early Modern and Present-day English is the rate of occurrence of negative raising, whereby the negative element in a subordinate clause, as in (36), moves to the main one, as in (37). (36) I think M. Wyat would no Englishman hurt, and this Enterprise cannot be done without the hurt and slaughter of both Parties; (1554 The Trial of Sir Nicholas Throckmorton I.55 c1; HC) (37) It does not seem to have any eye-lids, and therefore perhaps its eyes were so placed, that it might the better cleanse them with its fore-legs; (1665 Hooke, Micrographia 13.5, 211; HC) In both constructions the scope of the negation is the complement rather than the matrix verb itself, so that in (37) the meaning is ‘it seems not to have any eye-lids.’ Though negative raising has been on record since the OE period, its frequency increased steadily over time so that in Early Modern English it was fairly common, though not as much as in Present-day English, where it is said to be characteristic of colloquial styles (Denison 1998: 244).
7 The regulation of word order One of the major syntactic changes in the history of English is the word order shift that it underwent at some time between Old, Middle, and Early Modern English, whereby
39 Early Modern English: Syntax English ceased to have a general SOV word order with a verb-second (V2) constraint in declarative sentences and became the SVO language it is today. This change has been the subject of extensive research, and has been ascribed to a complex interplay of factors (cf. Seoane 2006). By late Middle English, the shift from verb-final to verb-medial word order was visible in the majority of subordinate clauses, and V2 was no longer predominant in declarative sentences. Early Modern English inherits from Middle English this strong tendency towards SVO word order; however, it also exhibits word order patterns which survive from Old English and which would cease to be possible in the 17th century. The following EModE examples have a word order which is disallowed today: (38) Then doo they vaunt themselues ouer the common multitude (1592 Nash, Pierce Penniless; Barber 1997: 191) (39) and only by theyr holle consent theyr citie and dominions were gouerned: (1531 Elyot, The Boke named The Governour; Barber 1997: 192) (40) Thys did I here hym saye (1500–70 Mowntayne, Autobiography 210; HC, Rissanen 1999: 266) (41) As we his subjects have in wonder found (1599 Shakespeare, King Henry V; Brinton and Arnovick 2011: 377) (42) I can thee thanke that thou canst such answeres deuise. But I perceyue thou doste me thoroughly knowe (1533 Udall, Roister Doister I.ii; HC, Rissanen 1999: 268) (43) Would I haue my flesh Torne by the publique hooke, these qualified hangmen Should be my company (1603 Jonson, Sejanus: His Fall II.iii; Rissanen 1999: 309) Example (38) exhibits VS or inverted word order after the initial adverbial then, a trace of the old V2 constraint. Such a word order pattern was also common, though not categorical, after also, here, now, so, there, therefore, thus, and yet. Jacobsson (1951) provides statistics for sentences beginning with some of these adverbs in our period, and his figures go from 34% of inverted patterns in the 16th century to only 7% in the 17th century. The verbs have, say, come and stand favor inversion much longer (Rissanen 1999: 265), but the regularization is nearly complete by the end of the EModE period. Example (39), on the contrary, has no inversion after the (semantically) restrictive adverbial only by theyr holle consent, a type of adverbial that, together with negative ones (such as never, nor), triggers mandatory inversion in Present-day English. The PDE rules for inversion after adverbials of this type were established over the course of the 17th century, first for negative adverbials (never, nor) and soon after for restrictive ones (seldom, rarely, little, etc.). Example (40) illustrates inversion of verb and subject (auxiliary – S – V) after a fronted direct object, another vestige of the old V2 constraint which would become progressively rarer in our period. The same fate would await examples (41) and (42). In (41) the subordinate clause retains the Old and Early Middle English verb-final
633
634
V Early Modern English (SOV) word order, and in (42) the verb phrase is split around the direct object, with verb-final order as well. Finally, example (43) is a conditional clause with inversion (auxiliary – S – V) and no conditional conjunction. Inverted conditional clauses are still possible today but are less frequent and are limited to the verbs be, have, and do (Were he to arrive earlier, please let him in; Rissanen 1999: 308). In the last three examples word order choices may have been influenced by metrical demands, since they are verse texts. The word order patterns illustrated in examples (38) to (43) are all optional, that is, in Early Modern English, like in Old and Middle English, there were no word order categorical rules but just strong tendencies of use. Word order variation in Early Modern English is controlled by an array of factors, such as metric and rhythmic considerations; the relative weight of the elements involved, that is, of the subject and object, also plays a role, in such a way that nominal subjects tend to occur postverbally more frequently than pronominal subjects, by virtue of their being longer. Stylistic factors have also been held responsible for word order variation, such as the connection between inversion and stylistic flourish (Rissanen 1999: 265). Finally, the author’s idiolect also determines the frequency of inversion in Early Modern English: thus, while William Roper, More, and Sydney prefer inversions after non-negative adverbials, Caxton’s use of inversion in this context amounts to only 8% of the cases (cf. Jacobsson 1951: 96–97). As already mentioned, the establishment ofwordorderrulesasweknowthemtoday tookplaceinthe17thcentury,withtheconsequent fixationofwordorderand a parallel increase in the use of alternative means of topicalization and order rearranging devices, such as the passive voice (cf. Seoane 2006).
8 Conclusions I hope to have shown, through the analysis of a selection of topics, that Early Modern English is a crucial period of transition between the still largely synthetic language found in Middle English and the new analytic and standardized English language we know today. This is a period when rules start to be fixed and when new changes, which are still ongoing today, begin to emerge. Though Early Modern English is no longer the neglected period of study it used to be and many powerful insights have shed light on its numerous developments, research is still needed in many areas, specially regarding the factors – linguistic, social – conditioning the changes it witnesses. Acknowledgments: I am grateful for generous financial support to the following institutions: the Spanish Ministry of Science and Innovation and the European Regional Development Fund (grant no. HUM2007-60706/FILO, FF12011-26693-CO2-01, and FF12011-26693-CO2-02); the Autonomous Government of Galicia (Directorate General of Scientific and Technological Promotion, grant no. 2008-047 the Directorate General for Research, Development and Innovation, INCITE grant no. 08PXIB204016PR).
9 References Allen, Cynthia L. 1986. Reconsidering the history of like. Journal of Linguistics 22: 375–409. Allen, Cynthia L. 1995. Case Marking and Reanalysis. Grammatical Relations from Old to Early Modern English. Oxford: Clarendon Press. Allen, Cynthia L. 1997. The development of an “impersonal” verb in Middle English: The case of behoove. In: Jacek Fisiak (ed.), Studies in Middle English Linguistics, 1–21. Berlin/New York: Mouton de Gruyter.
39 Early Modern English: Syntax Altenberg, Bengt. 1982. The Genitive v. the of-Construction. A Study of Syntactic Variation in 17th Century English. Malmo¨: CWK Gleerup. Barber, Charles. 1997. Early Modern English. Edinburgh: Edinburgh University Press. Blake, Norman F. 1983. Shakespeare’s Language: An Introduction. London: Macmillan. Brinton, Laurel J. 1988. The Development of English Aspectual Systems: Aspectualizers and Post-verbal Particles. Cambridge: Cambridge University Press. Brinton, Laurel J. 1996. Pragmatic Markers in English. Grammaticalization and Discourse Function. Berlin/New York: Mouton de Gruyter. Brinton, Laurel J. and Leslie K. Arnovick. 2011. The English Language. A Linguistic History. 2nd edn. Oxford: Oxford University Press. Brunner, Karl. 1955. Expanded verbal forms in Early Modern English. English Studies 26: 218–221. Cheshire, Jenny. 1998. English negation from an interactional perspective. In: Peter Trudgill and Jenny Cheshire (eds.), The Sociolinguistic Reader. Vol. I. 127–144. London: Arnold. Cuyckens, Hubert and Hendrik De Smet. 2007. For … to infinitives from Early to Late Modern English. In: Pe´rez-Guerra et al. (eds.), 77–103. Bern: Peter Lang. Denison, David. 1990. The Old English impersonals revived. In: Sylvia Adamson, Vivien A. Law, Nigel Vincent, and Susan Wright (eds.), Papers from the 5th International Conference on English Historical Linguistics. Cambridge 6–9 April 1987, 111–140. Amsterdam/Philadelphia: John Benjamins. Denison, David. 1993. English Historical Syntax. Harlow: Longman. Denison, David. 1998. Syntax. In: Suzanne Romaine (ed.), The Cambridge History of the English Language. Vol. IV. 1776–1997, 92–329. Cambridge: Cambridge University Press. Donner, Morton. 1986. The gerund in Middle English. English Studies 67: 394–400. Ellega˚rd, Alvar. 1953. The Auxiliary do: The Establishment and Regulation of its Use in English. Stockholm: Almqvist and Wiksell. Elmer, Willy. 1981. Diachronic Grammar. The History of Old and Middle English Subjectless Constructions. Tu¨bingen: Max Niemeyer. Elsness, Johan. 1994. On the progression of the progressive in Early Modern English. ICAME 18: 5–25. Elsness, Johan. 1997. Diachronic Grammar. The History of Old and Middle English Subjectless Constructions. Berlin/New York: Mouton de Gruyter. Fanego, Teresa. 1996a. The development of gerunds as objects of subject-control verbs in English (1400-1760). Diachronica 13: 29–62. Fanego, Teresa. 1996b. The gerund in Early Modern English: Evidence from the Helsinki Corpus. Folia Linguistica Historica 17: 97–152. Fanego, Teresa. 2004. On reanalysis and actualization in syntactic change: the rise and development of English verbal gerunds. Diachronica 21(1): 5–55. Fanego, Teresa. 2007. Drift and the development of sentential complements in British and American English from 1700 to the present day. In: Pe´rez-Guerra et al. (eds.), 161–235. Fischer, Olga. 1992. Syntax. In: Norman Blake (ed.), The Cambridge History of the English Language. Vol. II. 1066–1476, 207–408. Cambridge: Cambridge University Press. Fischer, Olga, Ans van Kemenade, Willem Koopman, and Wim van der Wurff. 2000. The Syntax of Early English. Cambridge: Cambridge University Press. Fischer, Olga and Frederike van der Leek. 1983. The demise of the Old English impersonal construction. Journal of Linguistics 19: 337–368. Fischer, Olga and Frederike van der Leek. 1987. A “case” for the Old English impersonal. In: Willem Koopman, Frederike van der Leek, Olga Fischer, and Roger Eaton (eds.), Explanation and Linguistic Change, 79–120. Amsterdam/Philadelphia: John Benjamins. Fischer, Olga and Wim van der Wurff. 2006. Syntax. In: Richard M. Hogg and David Denison (eds.), A History of the English Language, 109–198. Cambridge: Cambridge University Press. Fride´n, Georg. 1948. Studies on the Tenses of the English Verb from Chaucer to Shakespeare. Uppsala: Almqvist and Wiksell.
635
636
V Early Modern English Go¨rlach, Manfred. 1991. Introduction to Early Modern English. Cambridge: Cambridge University Press. Hogg, Richard M. (ed.). 1992. The Cambridge History of the English Language. Vol. I. The Beginnings to 1066. Cambridge: Cambridge University Press. Jacobsson, Bengt. 1951. Inversion in English with Special Reference to the Early Modern English Period. Uppsala: Almqvist and Wiksell. Jespersen, Otto. 1909–49. A Modern English Grammar on Historical Principles. 7 vols. Copenhagen: Ejnar Munksgaard. Kastovsky, Dieter. 1992. Semantics and Vocabulary. In: Hogg (ed.), 290–408. Kyto¨, Merja. 1997. Be/have + past participle: The choice of the auxiliary with intransitives from Late Middle to Modern English. In: Matti Rissanen, Merja Kyto¨, and Kirsi Heikkonen (eds.), English in Transition: Corpus-Based Studies in Linguistic Variation and Genre Styles, 17–85. Berlin/New York: Mouton de Gruyter. Lo´pez-Couso, Marı´a Jose´. 1996. On the history of methinks: From impersonal construction to fossilised expression. Folia Linguistica Historica XVII: 153–169. Loureiro-Porto, Lucı´a. 2010. A review of Early English impersonals: Evidence from necessity verbs. English Studies 91(6): 674–699. Mazzon, Gabriella. 2004. A History of English Negation. London: Longman. McWhorter, John. 2002. What happened to English? Diachronica 19(2): 217–272. Nehls, Dietrich. 1988. On the development of the grammatical category of verbal aspect in English. In: Josef Klegraf and Dietrich Nehls (eds.), Essays on the English Language and Applied Linguistics on the Occasion of Gerhard Nickel’s 60th Birthday, 173–198. Heidelberg: Julius Groos. Nu´n˜ez-Pertejo, Paloma. 2004. The Progressive in the History of English with Special Reference to the Early Modern English Period: A Corpus-based Study. Mu¨nchen: Lincom Europa. Ogura, Michiko. 1990. What has happened to “impersonal” constructions? Neuphilologische Mitteilungen 91: 31–55. Palander-Collin, Minna. 1999. Grammaticalization and Social Embedding: I THINK and METHINKS in Middle and Early Modern English. Helsinki: Socie´te´ Ne´ophilologique. ´ lvarez, Jorge Luis Bueno-Alonso, and EsperPe´rez-Guerra, Javier, Dolores Gonza´lez-A anza Rama-Martı´nez (eds.), “Of Varying Language and Opposing Creed”: New Insights into Late Modern English. Bern: Peter Lang. Rissanen, Matti, Merja Kyto¨, Leena Kahlas-Tarkka, Matti Kilpio¨, Saara Nevanlinna, Irma Taavitsainen, Terttu Nevalainen, and Helena Raumolin-Brunberg. 1991. The Helsinki Corpus of English Texts. In: ICAME Collection of English Language Corpora (CD-ROM), 2nd edn., Knut Hofland, Anne Lindebjerg, and Jørn Thunestvedt (eds.), The HIT Centre, University of Bergen, Norway. For manual, see http://khnt.hit.uib.no/icame/manuals/hc/index.htm Rissanen, Matti. 1999. Syntax. In: Roger Lass (ed.), The Cambridge History of the English Language. Vol. III. 1476–1776, 187–331. Cambridge: Cambridge University Press. Rosenbach, Anette. 2002. Genitive Variation in English. Conceptual Factors in Synchronic and Diachronic Studies. Berlin/New York: Mouton de Gruyter. Rosenbach, Anette, Dieter Stein, and Letizia Vezzosi. 2000. On the history of the s-genitive. In: Ricardo Bermu´dez-Otero, David Denison, Richard M. Hogg, and C. B. McCully (eds.), Generative Theory and Corpus Studies: A Dialogue from 10 ICEHL 183–210. Berlin/New York: Mouton de Gruyter. Rosenbach, Anette and Letizia Vezzosi. 2000. Genitive constructions in Early Modern English: New evidence from a corpus analysis. In: Rosanna Sornicola, Erich Poppe, and Ariel Shisha-Halevy (eds.), Stability, Variation and Change of Word-Order Patterns over Time, 285–307. Amsterdam/ Philadelphia: John Benjamins. Ryde´n, Mats. 1997. On the panchronic core meaning of the English progressive. In: Terttu Nevalainen and Leena Kahlas-Tarkka (eds.), To Explain the Present. Studies in the Changing English Language in Honour of Matti Rissanen, 419–429. Helsinki: Socie´te´ Neophilologique.
40 Early Modern English: Lexicon and semantics
637
Sapir, Edward. 1921. Language. An Introduction to the Study of Speech. New York: Harcourt, Brace and Co. Seoane, Elena. 2006. Information structure and word order: The passive as an information rearranging strategy. In: Ans van Kemenade and Bettelou Los (eds.), Handbook of the History of English, 360–391. Oxford: Blackwell. Smitterberg, Erik. 2005. The Progressive in 19th Century English: A Process of Integration. Amsterdam: Rodopi. Tajima, Matsuji. 1985. The Syntactic Development of the Gerund in Middle English. Tokyo: Nan’un-do. Traugott, Elizabeth Closs. 1992. Syntax. In: Hogg (ed.), 168–289. Visser, Frederikus Theodorus. 1963–73. A Historical Syntax of the English Language. 3 vols. Leiden: E. J. Brill. von Seefranz-Montag, Ariane. 1984. “Subjectless” constructions and syntactic change. In Jacek Fisiak (ed.), Historical Syntax, 521–553. Paris/The Hague: Mouton de Gruyter. Wischer, Ilse. 2000. Grammaticalization versus lexicalization: “Methinks” there is some confusion. In: Olga Fischer, Anette Rosenbach, and Dieter Stein (eds.), Pathways of Change. Grammaticalization in English, 355–370. Amsterdam/Philadelphia: John Benjamins.
Elena Seoane, Vigo (Spain)
40 Early Modern English: Lexicon and semantics 1 2 3 4 5 6
Introduction Resources for the study of Early Modern English lexis Lexicon Semantics Research issues References
Abstract As printing technology itself once did, cyberinfrastructure is changing research into the Early Modern English lexicon and semantics. The Online OED, Early English Books Online, Lexicons of Early Modern English, The Textbase of Early Tudor English, and other resources now enable us to chart the growth of English in great detail. The mother tongue remained small, well under 10,000 words, until the 17th century. Printed books, however, by saving and disseminating learned and technical words, expanded available vocabulary by 75% from 1500 to 1600. Hundreds of glossaries and dictionaries printed word-entries that mapped English terms to each other and to other tongues and stimulated interest in semantics. Nouns and verbs were no longer assumed to be names for things and actions, as the famous Lily-Colet grammar taught, but (especially influenced by John Locke in 1690) became pointers to ideas in individual minds. Under the early Tudors, English was widely maligned as a minor tongue lacking the vocabulary and the sophistication of ancient and modern languages. Researchers can now chart a full account of how the English made their own tongue competitive. Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 637–652
638
V Early Modern English
1 Introduction Speakers of Early Modern English (EModE) lived through a language-technology revolution comparable to what we have experienced since 1964, when IBM introduced its Magnetic Tape/Selectric Typewriter, an early industrial forerunner of the word processor (see http://www-03.ibm.com/ibm/history/exhibits/modelb/modelb_office.html, last accessed 14 September 2011). Printing enabled early Tudor writers to archive and disseminate their new words and senses, leading to the making of widely-available bilingual and monolingual dictionaries and grammars. This in turn contributed to lexico-semantic enrichment, strengthened the case for standardization, and spawned revisionist spelling systems and universal-language schemes (Cohen 1977; Slaughter 1982). The EModE period, for these reasons, was a time of more than just routine semantic change. It witnessed a sizable increase in vocabulary and a new interest in semantics itself. Speakers, readers, and writers had to pay attention to word meaning, thanks to the growth of word-loans in works translated from other languages, of socalled “terms of art” created by professional, technical, and social groups with a special expertise, and of simple zero-derivation, an innovative technique favored by Shakespeare, whereby a new word could be made almost transparently from an existing term by altering its part of speech. Anyone could create new words and senses, and few reading or listening to them were alert enough to detect, or object to, these lexical sports. By 1656, when Thomas Blount published Glossographia, a hard-word lexicon, readers faced a schizophrenic, split lexicon: a mother tongue or “common core” (Nevalainen 2006: 47) anchored in the ordinary talk of illiterate and literate alike, and by the rites and scriptures of the Church of England; and a blizzard of invented, borrowed, and transformed words to which no guide existed. For the first time, also, the English lexicon suffered few enduring, unredeemable losses. Given the archival value of printed books, words that passed out of usage at that time are still available. Our knowledge of the state of the EModE lexicon far surpasses that of most people alive in that period. Educated men and women in Tudor England knew more about and respected Latin better than their own tongue. Sir Thomas Elyot’s Dictionary (1538) was extended by Thomas Cooper in many re-editions up to 1584 before it was eclipsed by Latin lexicons from the likes of Thomas Thomas (1587) and John Rider (1589), which competed more effectively than had earlier dictionaries by Richard Huloet (1552) and John Baret (1574). The average English nobleman in 1617 could have purchased remarkably full lexicons of Latin, French, Italian, and Spanish, as well as John Minsheu’s (1617) astonishing Ductor in Linguas, an etymological lexicon of English that cited eleven languages. If not always despised, English was regarded as a severely understocked old tongue, dependent on absorbing words from many other languages if it was to remain viable. Even defenders of English, like Sir Philip Sidney (1598), insisted that it ought not be taught, despite Merchant Taylors’ schoolmaster Richard Mulcaster’s urgings to the contrary. Those who praised English to the skies, like Richard Carew in “The Excellency of the English Tongue” (1614), proposed as an advantage that “the most parte of our wordes […] are Monasillables, and soe the fewer in tale, and the sooner reduced to memorye” (Go¨rlach 1991: 241). It is little wonder that Shakespeare confessed himself, as a vernacular playwright, to be one who “sold cheap what is most dear” (Sonnet 110) in comparison to an envied rival poet, likely George Chapman, the translator of Homer’s Odyssey.
40 Early Modern English: Lexicon and semantics The Renaissance produced some astounding feats of lexico-semantic analysis, but they had little to do with the English language. Robert Estienne’s great dictionary of classical Latin, and his son’s comparable achievement for Greek, belonged to France; the Accademia della Crusca produced the first great dictionary of a vernacular language, Italian, in its Vocabolario of 1612; and the Dictionnaire de l’Acade´mie Franc¸aise, first published in 1694, gave dignity to French. Considine (2008) has recently discussed such major works. The great English literature of the mid-Renaissance in England, works such as Spenser’s The Faerie Queene, Shakespeare’s tragedies, and the King James Bible, however, used a national language that was lexicographically unrecorded except in bilingual lexicons. Most grammar schools taught from the partly translated Lily-Colet Latin grammar (1549). The first monolingual textbook of spelling, dates, and hard words, intended for pre-grammar-school students and uneducated adults, was Edmund Coote’s The English School-maister (1997 [1596]): it eventually became a standard work. Using his own non-standard spelling system, William Bullokar published books on English grammar in the 1580s; Alexander Gill’s English grammar, Logonomia Anglica (1619–21), was entirely in Latin, aside from English-language examples in his own non-standard spelling. However sophisticated may have been the linguistic ideas of a very few, most EModE authors in the first half of our period gave little credit to the vernacular. Jones (1953: 3–167) describes how contemporaries viewed English as uneloquent, inadequate, and misspelled.
2 Resources for the study of Early Modern English lexis Analysis of this period’s vocabulary and word-meaning began seriously with the publishing of the complete Oxford English Dictionary (OED) (and first supplement) in 1933, and the start of the ongoing University Microfilms Microfilm Project in 1938, which aimed to make available images of all printed books in the Short-Title Catalogue (STC) (Pollard and Redgrave 1976–91; Wing 1972–88 and their supplements). When the planned Early Modern English Dictionary (EMED) project lapsed in the late 1930s, the OED became the default source for the period’s lexico-semantic data. Richard Bailey published a large card-file of EMED additions to and antedatings of OED word-forms and senses in 1978. Ju¨rgen Scha¨fer (1980) then made two significant contributions to the study of the period language by demonstrating the bias inherent in OED citation selection for the Early Modern period and by supplementing OED with data on 5,000 words found in English printed glossaries from 1526 to 1640 (Scha¨fer 1989). Two other great national dictionaries for old Scots (Aitken and Craigie 1931– 2002) and Welsh (Thomas et al. 1967–2002) have since appeared. In 2006, the University of Toronto Press and Libraries also brought out my online database, Lexicons of Early Modern English (LEME) (Lancashire [ed.] 2006–), which now includes 575,000 word-entries from 166 dictionaries and glossaries of the period. LEME replaces a prototype Web textbase, Early Modern English Dictionaries Database (EMEDD) (Lancashire [ed.] 1999), which I developed online from 1996 to 1999. In 2003, ProQuest introduced Early English Books Online (EEBO; Chadwyck-Healey 2003–2011), which in its first five years has made available image reproductions of about 122,000 works from STC, Wing, and their supplements. STC and its collections have been supplemented by two magnificent research projects by Robin Alston (1965): Bibliography of the English Language […] to 1800, a magnum opus still in progress, which
639
640
V Early Modern English identifies, classifies, and extends the body of language texts that EEBO now covers; and 365 facsimile reprints of such books for Scolar Press’s series English Linguistics, 1500– 1800. In 1999 the University of Michigan and ProQuest entered, with member institutions worldwide, into the Text Creation Partnership (TCP) to digitize 25,000 works from these collections. As of September 2011, EEBO-TCP (Welzenbach 1999–) has released 32,957 searchable full-texts. It is possible to search the spelling-list from these data now, but the number of English word-forms or lemmas that these spellings represent is not known. The EEBO-TCP index treats uppercase and lowercase forms separately and does not distinguish between English and other languages. For example, querying the index for spellings of “definition” retrieves several dozen spellings, including “definition” (8,244 occurrences), “definitions” (1,575), “Definition” (4,215), “definitione” (324), “definitiones” (125), “Definitione” (40), “Definitiones” (33), “definicion” (40), “DEFINITION” (46), and “DEFINITIONS” (8). Lemmatizing of all these data, now underway by several groups, will enable the OED to assimilate the many new words, senses, and antedatings they document. However, manual lemmatizing is an immense task, and the error-rate of semi-automatic lemmatization probably still remains too high at the present time. Modern editions of period dictionaries, diachronic corpora, and analytic dictionaries of the period by present-day lexicographers add substantially to a researcher’s lexicosemantic data. The Early English Text Society has turned out editions of Catholicon Anglicum (1483; Herttage [ed.] 1881), Promptorium Parvulorum (1440–99; Mayhew [ed.] 1908), and Peter Levins’s Manipulus Vocabulorum (1570; Wheatley [ed.] 1867). Other stand-alone editions include Vocabularium Saxonicum by Laurence Nowell (c.1565; Marckwardt [ed.] 1952), Dictionarie of the Vulgar Russe Tongue by Mark Ridley (c.1594–99; Stone [ed.] 1996), L’e´claircissement de la langue franc¸aise by John Palsgrave (1530; Baddeley [ed.] 2003), Libellus de re herbaria novus by William Turner (1538; Ryde´n et al. [eds.] 1999), and Table Alphabetical by Robert Cawdrey (1604). (For a complete listing of lexicons from 1480–1702, see http://leme.library.utoronto. ca/menu/menuLexicon.cfm.) Contemporary essays on English, and prefaces and introductions to dictionaries, grammars, and language textbooks, which Manfred Go¨rlach (1991) surveys, are a valuable barometer of language consciousness in the period. The Helsinki Corpus (Rissanen et al. 1991; see also Kyto¨ 1996; Rissanen et al. 1993) and the Corpus of Early English Correspondence (Nevalainen et al. 1998; see also Nurmi 1998) focus on grammatical and syntactic features but also encode texts (and thus their vocabulary) for various features that bear on semantics. Waite’s The Textbase of Early Tudor English (2008) is a TEI-encoded Web collection of some 260,000 lines of verse, 1485–1550, that plans to develop a rhyme index and an index of early Tudor spellings.
3 Lexicon The Early Modern period begins with the introduction of printing to England. This technology created an archive for English. Tudor lexicographers and translators froze new words in print, where they could survive long after anyone stopped uttering them or writing them down. Stein (1985) showed why Robert Cawdrey’s (1604) little hard-word lexicon was far from being the first English dictionary – as so many claimed it was for much too long – when she traced a line of English lexicographical works well back into the Old English period. The growth of the hard-word glossary – which Scha¨fer
40 Early Modern English: Lexicon and semantics (1989) documents from 1526 on – testifies to such word-loans, mostly from Latin, technical “terms of art” from trade and craft guilds and from professions like law, pharmacology, and medicine. The 16th and 17th centuries so expanded this mass of new, unfamiliar words that it became a second English, rivalling the mother tongue that people used unselfconsciously and thought little about. In Ordered Profusion, Finkenstaedt et al. (1973) used the first-occurrence dates of OED headwords in their Chronological English Dictionary (CED) (Finkenstaedt et al. 1970) to identify 1560–1660 as the peak period of vocabulary growth. McDermott (2002) linked the CED rate of lexical expansion with the number of books printed yearby-year according to the STC (the latter calculated by Bell and Barnard 1992). Printed books are a persistent quantitative measure in the period, and books created the language’s lexical archive. The most lexically rich works are dictionaries. Totals of new CED words increase over a century from 1490–1509 to 1590–1609 by 75%. That astonishing rate of growth resembles the increase of lexicon production more than the increase of all STC books. By 1500, English had just over 36,100 different wordforms, and 64,300 by 1600. This is an increase of 78%, or 3% more than CED numbers from 1973. The OED Online (Simpson [ed.] 2000–), appearing forty years after the publication of the CED, now allows us to recalculate these data. John Simpson, the editor of the OED, described a simple strategy to select, for instance, the “available vocabulary” between 1650 and 1660: “go to the Advanced search page (leaving ‘entries’ selected, rather than ‘quotations’), search for 500–1660 in the first search field (by ‘quotation date’), AND 1650–2008 in the second search field (again by ‘quotation date’)” (p.c. of March 13, 2008). His result includes affixes and combining forms, which are not awarded “first dates” by the database, and so yields a slightly higher count. My preliminary count of lemmas in LEME uses over 95% non-OED texts but indicates an increase of 75%, like the CED. Ongoing large-scale antedatings of OED headwords first dated in the Early Modern period indicate that the actual rate of increase of Early Modern English vocabulary may be lower. More and more, we are finding that words occurred earlier than the OED record shows. Bailey’s (1978) collection of additions and antedatings to English vocabulary, 1475–1700, has 4,400 entries, of which about 2,900 included antedatings (based on a hand-count of antedatings on twelve pages of the 330-page “Additions and Antedatings.”) Scha¨fer’s (1980) study of OED sources revealed that the early Oxford lexicographers had relied overly on citations from major authors, especially Shakespeare, and under-represented the more lexically inventive, such as Thomas Nashe. In 1989 Scha¨fer also demonstrated that OED overlooked information in hard-word glossaries from 1526 up to the mid-17th century. The 5,000 word-entries in his Early Modern English Lexicography (Scha¨fer 1989) make about 2,600 antedatings to the OED (based on a hand-count of antedatings on twelve pages of the 177-page “Additions and Corrections”). Thus, by 1989 just two scholars had antedated 5,500 OED Early Modern headwords, out of a total of 64,000. We can also infer that the growth of printing, as previously mentioned, archives new words in a way not possible in the Middle English period. The word-list from the STC books recently digitized by EEBO-TCP (Welzenbach 1999–) will also have an impact. This process of antedating – up to 50% of OED headwords – is spreading vocabulary growth out more evenly through the Tudor period and may diminish the “acceleration” effect. The total archival and mother
641
642
V Early Modern English lexicons of English will still increase, but their growth-rate in the Tudor period may flatten somewhat unless the OED’s new sources produce a balancing number of neologisms. The evidence for these may already exist because, as John Considine (who has worked at the OED) tells me, its office “has large files of ‘pre-1800 notins’, i.e. words known from sources of before 1800 which are not registered in the dictionary because their sense is obscure, or they appear to have been imperfectly naturalized, or they’re just very rare” (p.c. December 10, 2008). Was it just hard-word English that expanded, or did the common core also? The schoolmaster of the Merchant Taylors’ School in London, Richard Mulcaster (1582) lists 8,143 different English words as the basis for a monolingual English dictionary of “our ordinarie speche” that, he hopes, someone else will compile. He allows for having missed some vocabulary items, but he also conflates the mother tongue, which he describes as the native vocabulary that everyone uses naturally, with newly enfranchised words (that is, word-loans) that people must learn before they can use. The English words employed in bilingual lexicons beginning with John Palsgrave’s English–French L’e´claircissement (1530) may be expected to come from the mother tongue, words that required no explanation. For example, John Withals, whose pocket lexicon (1553) was popular throughout the 16th century, wrote as if he expected his grammarschool students to have a basic English vocabulary of fewer than 4,000 words. Even the written subset of Shakespeare’s vocabulary over twenty years, from 1590 to 1613, was only twice the size of Mulcaster’s list. Alfred Hart (1943) estimated it to be 17,677 word-forms. (The nearly 40,000 entries in the Riverside concordance of Shakespeare [Spevack 1973] can be used to exaggerate Shakespeare’s personal working lexicon. The concordance includes variant spellings, and plural forms of words, among its headwords.) There are thus acceptable grounds for a small EModE mother tongue that had many fewer words than the 64,300 lemmas recorded by the OED Online (Simpson [ed.] 2000–). By 1600, the population of England was divided among the country (90%), greater London (5%) (Sheppard 1998: 363), and the other provincial cities (5%; Guy 1988: 34–35). That is, nine of every ten persons lived in places that had no book-sellers. Existing data tell us that some 80% of all Elizabethan men were illiterate, and 95% of women. Even by the 1640s, 70% of men in rural England could not sign their name, in contrast with 22% of Londoners (Guy 1988: 417), although Thomas (1986: 103) expresses doubt that the number of people who could sign their name is a reliable marker of literacy. He believes the literacy rate was much higher than existing data indicate. Country folk as a rule, however, may have had only one book, the English Bible. For example, William Tyndale had used the common mother tongue so that scripture could be understood when it was read aloud at home and church. The plowboy-reader that he had in mind did not need any help to understand what the Bible meant; the vocabulary of the New Testament was modest. If the mother tongue, the English language used by 90% of the English people, had grown markedly, ordinary people would have needed general-purpose monolingual English dictionaries, but none was published until the early 18th century. John Kersey first brought out what he described as a general dictionary of English in 1702. Few before that time documented the mother tongue that parents passed on to children in daily usage, although teachers of English spelling did so. John Evans (1621), in his The Palace of Profitable Pleasure, listed “all English words” alphabetically, each one
40 Early Modern English: Lexicon and semantics divided into syllables: the number of words needed to read the Bible and other books, he says, adds up to 6,665. David Crystal (2004: 317) finds 8,000 lexemes in the King James Bible, a number that generally confirms Mulcaster’s estimate forty years before. What grew were occupational sociolects, the specialized vocabularies of city folk, especially the 10% of the population that lived in London (which had tripled in size by the late 16th century) and the provincial cities. A mass of loan-words from the Latin of the educated, and the French, Italian, and Spanish of the merchants, combined with “terms of art” in burgeoning fields of knowledge such as medicine (McConchie 1997), law, and pharmacopoeia (mainly herbs) to form a second tongue, rival to the mother tongue. By the Stuart period, hard words had grown to such an extent that hard-word lexicons by Robert Cawdrey (2007 [1604]), John Bullokar (1611), and Henry Cockeram (1623) had to be printed. By the late 17th century, lexicographers like Thomas Blount (1656) and Elisha Coles (1676) were producing full-length dictionaries that served only hard words. The internationalist diplomacy of James I, the religious and political turmoil of the civil war, and the rising sciences that would join into the Royal Society in 1660, among other things, explosively accelerated the growth of this second tongue in the post-Tudor world. The English did not learn hard words and terms of art from the publication of lexicons but rather because of professional and social communities in London that created conditions for specialized sociolects. Vocabulary expertise gave people a path to wealth in a land where population had doubled and resources had not. The growth of population, the failure of commodities to keep up to need, the declining purchasing power of the wages of most English laborers, the centralization of the crown and its income in London, and the huge development of new media – the printing press and the playhouse – led ingenious people to expand their professional vocabularies. Most OED sources for the Early Modern period are books published in London for an emerging knowledge-based elite, which was at most 10% of the population. Rural and poor speakers could neither use nor make money from specialized vocabularies, but urban professionals could, and their need for money was the root of the growth of a second tongue.
4 Semantics In the first half of the Early Modern period, most literate English speakers had only the most rudimentary understanding of word meaning and semantics. Caxton’s observations on English nearly exhaust what the literate English knew about the semantics of their own speech: change dominated it, whether for dialectal reasons or personal taste. Loo what sholde a man in thyse dayes now wryte. egges or eyren/ certaynly it is harde to playse euery man / by cause of dyuersite & chaunge of langage. For in these dayes euery man that is in ony reputacyon in his countre. wyll vtter his commynycacyon and maters in suche maners & termes / that fewe men shall vnderstonde theym (1490 Caxton’s Preface to Virgil’s Eneydos A1v).
Renaissance scholars writing and conversing in Latin, especially on the Continent, knew much more about semantics than Caxton did (Salmon 1990), of course, but they did not favor English as a subject of study. Early in the reign of Henry VIII, Erasmus (who
643
644
V Early Modern English never learned English during his half a dozen years in England) described English as a tongue of monosyllables – a trait that the Cornishman Richard Carew, a century later, boasted about – and said that the vernacular he had heard used in taverns by men who were hawking merchandise sounded like barking (Giese 1937: 11). Once the lexicon markedly grew, thanks to printing and other factors (such as the translation of works from other languages, and the growth of bilingual dictionaries), even the educated English speaker only very gradually woke up to the problematic meaning of words. English words in the Tudor period were generally understood to have just two semantic traits: they were easy or hard, and they denoted something in the external world. Early Modern English glossographers, accordingly, had one main function: to offer, for any hard English word, easier synonyms and corresponding expressions. Monolingual glossaries explained those English words or terms of art which native speakers had to learn by easy terms they knew naturally (as Mulcaster 1582: 166–168 says). Bilingual lexicons gave easy English words for foreign-language words. The dozens of these dictionaries and glossaries that reached print did not document meaning in the common understanding we have of it today. The modern concept of a lexical definition was not to be found in them, outside occasional texts about mathematics, geometry, and the like. Early lexicographers often used synonyms, ornamented by anecdotes, for the postlemmatic explanation field (cf. Stein 1986; Lancashire 2006). Samuel Johnson (1979 [1755]) did not recognize the lexical use of the word “definition”. Native speakers who recognized that words had different senses did not seek to explain these differences carefully. The Lily-Colet grammar book [1549] that the English Crown assigned for use in schools in the reign of Edward VI, and that lasted for more than a century, asserted that nouns were names of things in the world (cf. Anderson 1996; Lancashire 2002). When Shakespeare’s Juliet asked “What’s in a name?” and said that “a rose / By any other word would smell as sweet” (II.ii.43–44), she assented to a theory of semantics that was concept-vacant. Francis Bacon alludes to the prevalence of this belief when he said that “the first distemper of learning, [is] when men studie words, and not matter” (Go¨rlach 1991: 247). This lexical model did not link words with concepts or ideas, a theory put forward in the late 15th century by the Italian humanist Lorenzo Valla, who believed that words, themselves changing over time like material things, could signify themselves (Wasno 1987: 109–10). John Locke in his Essay Concerning Human Understanding (1690) persuasively established this semantic notion in England much later. He wrote that “Words then are made to be signs of our Ideas”, which were originally (and still are, often enough) “sensible Ideas” (Go¨rlach 1991: 401–402; cf. Cohen 1977). We owe this shift in metasemantic thought to Locke and philosophy, and it was to have enormous influence in centuries to come. Once the human mind could be said to have, within it, everything in the world merely by reason of naming those things, people could shake off the constraints that tied their words directly to the world. It became possible to build inner mental worlds as models for how to shape history and nature. Ironically, the founding of the Royal Society in 1660 did not entirely support a scientific method. When John Wilkins in 1668 published an analysis of English semantics that charted connotation and thesaural relationships among words, he was remaking language as an intellectual system, independent of empirical observation. This followed from the myth of the Tower of Babel, when God punished overreaching man by making
40 Early Modern English: Lexicon and semantics him speak different tongues. Before Babel, everyone was thought to share the same Adamic tongue. Acting as God’s regent on earth, the first man gave names to all the animals. Those names could hardly be arbitrary unless God acted without reason. After Wilkins and Locke, words and the ideas they signified were synonymous and inherently subjective. A key to exercising power over human beings and their world lay in forging a semantically self-consistent inner vision that subjected words to working concepts, and in then getting others to accept that semantic map. The easy name-to-thing denotation that the Lily and Colet (1549) grammar book assumes became problematic when English speakers grew more interested in analyzing English words rather than, in a straightforward way, just assigning to them synonymous terms. Elyot (1542) expanded his scope from words to phrases, and he added “true definitions of all syckenesses and kyndes of maladyes” (sig. a2v). Elyot merged some definitions of things into a dictionary of words. Richard Huloet (1552: sig. [x]2v) observed in the preface to his English-Latin lexicon that, in Britain, English words appeared in Latin text, and Latin words in English text. Both languages, then, deserved equitable treatment. Thomas Wilson (1553), an early defender of English, wrote that “Some seke so farre for outlandishe Englishe, that thei forget altogether their mothers language” (p2r). Richard Mulcaster (1582) plainly observed that “verie manie men, being excellentlie well learned in foren speche, can hardlie discern what theie haue at home” (x4r). “Our naturall tung,” he said, “cummeth on vs by hudle [in confusion]” (x4v), and yet “the word being knowen, which implyeth the propertie the thing is half known, whose propertie is emplyed” (y1r). Thomas Thomas (1587: 4) five years later cited approvingly Plato’s dictum that whoever knows the names of things will know the things also. Robert Cawdrey (2007 [1604]: a3r) copied Wilson’s above remark (Crystal 2004: 291). Francis Bacon (1605) in his Aduancement of Learning argued that “wordes, are the tokens currant and accepted for conceits” (II, 60), in this way associating semantics with thought. When John Bullokar (1616), in his hard-word dictionary, claims to “open the signification of such [strange] words” (a3v), some readers understood “signification” as more than acting as signs for things. For example, individuals were confident of their personal interpretations of scripture, a practice that seeded puritan sects on both sides of the Atlantic, sects that promoted widespread religious dissent, a civil war, and the Interregnum. The growth of the core and hard-word lexicons highlighted semantic issues. When synonyms for a single mother-tongue term multiplied in word-loans from different sources, copiousness resulted, and so did doubts about usage. John Florio (1598) enjoyed multiplying terms and senses and praised English, given the “manie-folde Englishes of manie wordes” in his explanations, for out-vying Italian itself (b1v). Under what circumstance was one synonym used rather than another? Another issue was ambiguity, when a single word acquired two or more senses, commonly because it developed a narrower sense in a professional or trade register. How could a speaker resolve puzzlement when a common term was used in an unexpected way because of its professional or technical context? Edmund Coote (1997 [1596]), for one, solved this problem by ignoring the common sense (l1r). The monosyllabic mother tongue of small size that Coote taught did not create many such problems, but when English almost doubled in a century because of word-loans and terms of art, readers and speakers could indeed become “gravelled” (Blount 1656: A2r) in understanding their own tongue.
645
646
V Early Modern English The prefaces of Early Modern dictionaries thus supply milestones in the English speaker’s gradually growing awareness of semantic issues. First was Sir Thomas Elyot’s shocked recognition that natural understanding of mother-tongue words was sometimes wrong. In 1538, he expressed dismay at one class of such errors, the misidentification of herbs by ignorant medical practitioners (Elyot 1538). Elyot had to rely on England’s first great herbalist, William Turner, to untangle the relationship between names and herbs. Until John Rider (1589) enumerated senses in his English-Latin dictionary, English lexicographers had separated synonyms and senses indiscriminately by commas. By 1658, Edward Phillips proposed “to distinguish the terms, several derivations, differences, definitions, interpretations, proper significations of the words of our Tongue” (Phillips 1658: a3v), even if he did not, in practice, know what all of these were, having plagiarized most of his dictionary from Blount. After the foundation of the Royal Society, lexicographers such as John Worlidge, John Ray, and Stephen Skinner made substantial contributions to neglected fields of English vocabulary. Worlidge (1669) analyzed dialectal variation in husbandry – that is, farming terminology. Ray (1674) surveyed British dialectal variation, north and south, and compiled word-lists for birds and fishes. The first glossary of Chaucer’s medieval words had come out in 1602: this showed that historical time as well as profession or trade could give rise to semantic ambiguity. This recognition led to the making of etymological lexicons (Liberman 1998). Stephen Skinner (1671) corrected etymologies in John Minsheu’s 1617 derivative, often erroneous Ductor in Linguas, drawing on William Somner’s 1659 dictionary of Anglo-Saxon, the product of almost a century of scholarship by half a dozen people.
5 Research issues Analyzing Early Modern English semantic change satisfactorily, a principal interest of present-day researchers, is, as Manfred Go¨rlach (1991: 200) says, “a difficult and largely unsolved problem”. Researchers have successfully classified specific instances of semantic change as “expansions, reductions and transfers” (Go¨rlach 1991: 207), or as generalization, specialization, pejoration, and amelioration (Nevalainen 1999: 433–434). However, until we can interrogate the OED database to search for word-entries that document the addition and the loss of senses, year by year, research will be partial and somewhat impressionistic. At present, the OED Online (Simpson [ed.] 2000–) advanced search function does not permit queries on senses. We could benefit from a count of lemmata that have undergone semantic change within successive, short periods of time. Then we could analyze each change for its possible causes and effects. Did one or more synonyms appear in the preceding decades and replace the headword, which was then available for semantic reassignment? Were some translations especially productive of sense-shifts? Did social, technological, religious, or intellectual developments highlight gaps in the lexicon that had to be filled? The OED does not intend to answer some of these questions, but other publications, such as Kay et al.’s (2009) Historical Thesaurus of English (now part of the OED), may help us do so. To understand semantic change is in part to comprehend to whom the native speakers of the language at the time ceded the right to influence how they used words. Crown patronage – in publishing dictionaries and reference works such as herbals, in supporting antiquarian studies in archaic forms of English, in enabling or censoring debate over
40 Early Modern English: Lexicon and semantics controversial subjects – had an impact on new-word formation and storage. Sir Thomas Elyot went so far as to credit his 1538 Latin-English dictionary to Henry VIII because of the king’s enthusiasm for the project and willingness to open up the royal library to his research. The English Crown normally expressed its patronage by allowing lexicographers to dedicate works directly to them or to their chief ministers (Williams 1962). The state and the church, whenever revising educational standards, liturgies, or the translation of the Bible, nudged forward new intellectual models or paradigms of language itself into the public eye. We should consider asking who has political power and money before assigning a cause for given semantic changes. Did the avowed preferences of William Cecil, Lord Burghley, for Latin over English, or of herbals and lexicons over poems and plays, influence Tudor English? After all, he brought Elizabeth to the throne, and he guarded her for fifty years. How did the assumption of James I in 1603, the arrival of thousands of Scots in London, his choice of Francis Bacon as Chancellor, and his humiliation of John Cowell and his legal lexicon, The Interpreter (1607), affect English, which had been taken for granted for more than a century? James saw himself as a European monarch: he sought alliances for his children across the English Channel. It was possible to buy two huge bilingual dictionaries of French (Randle Cotgrave 1530) and Italian (John Florio 1598), both dedicated to royals, in 1611, when English itself had to make do with the non-headword position in these massive language works. John Minsheu (1617) had English headwords in his Ductor in Linguas, but they were remarkably limited, a few thousand, and overwhelmed by the ten other languages whose etymological relations to English the lexicon’s word-entries highlighted. The OED Online (Simpson [ed.] 2000–) has unrivalled scholarship on Early Modern English semantics, but neither the technology to extract it (yet), nor sufficiently representative illustrative contemporary quotations. If we link authors of quotations in the OED with biographical entries about potential authors in the Oxford Dictionary of National Biography (Matthew et al. [eds.] 2004–08), we find a discontinuity. About 25 quotations by William Cecil, 50 by Elizabeth I, and 55 by James I appear in the OED, but over 5,600 by Shakespeare. In the tradition of Samuel Johnson, the OED originally founded its base of quotations on excellent speakers and writers of literary English, not on men and women who exerted a powerful influence on all aspects of the nation, and in particular its language. Yet we have a substantial manuscript literature written by Cecil and (especially) James, and no mean remains from Elizabeth. The same is true of dictionary writers. Their neglect, as a group, is so marked that less than 5% of the word-entries in Lexicons of Early Modern English (LEME) (Lancashire [ed.] 2006–) are (independently) cited by the Online OED. Is it possible that the semantics of Early Modern English appears so intractable a problem to language historians because present-day researchers analyze its language separately from the people who spoke it, and the world they describe? A hard English headword in a glossary of the time gave glossographers an opportunity to write a definition, but they seldom did so. They often preferred anecdotes, personal reflections, and information about the thing that the word denoted. The perception of lexical senses, unstandardized at this time, was individuated and contextualized in ways important to the writer and speaker. John Rastell’s early law lexicon gives earlier instances of a glossographer whose explanations closely mirror the living context in which they were used. His Exposiciones Terminorum Legum Anglorum (1523–24) “glosses ‘derke termys’ in ways that serve royal power” (Lancashire 2006: 13). Much later, Thomas
647
648
V Early Modern English Blount’s (1656) entry on “Landskip” (‘landscape’) sharply contrasts with the current definition for this sense of the word in the OED (s.v. landscape, def. 1b ‘The background of scenery in a portrait or figure-painting’, 1656–76): Landskip ( Belg.) Parergon, Paisage or By-work, which is an expressing of the Land, by Hills, Woods, Castles, Valleys, Rivers, Cities, &c. as far as may be shewed in our Horizon. All that which in a Picture is not of the body or argument thereof is Landskip, Parergon, or by-work. As in the Table of our Saviors passion, the picture of Christ upon the Rood (which is the proper English word for Cross) the two theeves, the blessed Virgin Mary, and St. John, are the Argument: But the City Jerusalem, the Country about, the clouds, and the like, are Landskip. El. Ar.
The OED Online gives Blount’s entry as its earliest citation in this sense, but Blount copies the glossary entry for “LANDSKEP” (Dd4v) almost verbatim from Edmund Bolton’s earlier Elements of Armories (1610). LEME finds a brief explanation of this sense of the word in both Cotgrave (1611) and Florio (1598). However, Bolton and Blount were Roman Catholics and associated landscape with religious images, the background for Christ’s crucifixion. The sense of this term was, as it were, marked by individual religious belief. When John Milton’s nephew Edward Phillips (1658), in his New World, plagiarized Blount’s dictionary, he copied the entry on “Landskip” but, like the OED, excised the reference to the religious image that, for Bolton and Blount, whose faith worshipped images, was relevant to the word’s signification. Milton and Phillips despised the Roman Catholic hegemony. What does this small instance of semantic change tell us, if not that classifying early senses by type – for example, as reductions – is bound to be unsatisfying? The intentions and context of the contemporary English speaker and writer, elusive though they may be, may be better explicators of historical semantic change. Acknowledgments: The research for this chapter was done at my Lexical Analysis Laboratory, Room 7061, Robarts Library, a facility that I owe to the support of Geoffrey Rockwell’s TAPoR network, IBM Canada, the Canada Foundation for Innovation, and the University of Toronto Libraries.
6 References Aitken, A. J. and William A. Craigie (eds.). 1931–2002. A Dictionary of the Older Scottish Tongue, from the Twelfth Century to the End of the Seventeenth. 12 vols. London/Chicago: Oxford University Press and University of Chicago Press. Alston, R. C. 1965. A Bibliography of the English Language from the Invention of Printing to the Year 1800. Leeds: E. J. Arnold. Anderson, Judith. 1996. Words that Matter: Linguistic Perception in Renaissance English. Stanford: Stanford University Press. Bacon, Francis. 1605. The Twoo Bookes of Francis Bacon. Of the Proficience and Aduancement of Learning, Diuine and Humane. London: Henry Tomes. Bailey, Richard W. (ed.). 1978. Early Modern English: Additions and Antedatings to the Record of English Vocabulary 1475–1700. New York: Georg Olms Verlag. Baret, John. 1574. An Aluearie or Triple Dictionarie, in Englishe, Latin, and French. London: Henry Denham. Bell, Maureen and John Barnard. 1992. Provisional count of STC titles, 1475–1640. Publishing History 31: 48–64.
40 Early Modern English: Lexicon and semantics Blount, Thomas. 1656. Glossographia. London: Thomas Newcomb for Humphrey Moseley and George Sawbridge. Bolton, Edmund. 1610. The Elements of Armories. London: George Eld. Bullokar, John. 1616. An English Expositor: Teaching the Interpretation of the Hardest Words in our Language. London: John Legatt. Carew, Richard. 1614. The excellencie of the English tongue. In: William Camden (ed.), Remaines, concerning Britaine, 36–44. London: John Legatt for Simon Waterson. Cawdrey, Robert. 1604. A Table Alphabeticall, Conteyning and Teaching the Understanding of Hard Usuall English Wordes, Borrowed from the Hebrew, Greeke, Latine, or French, & c. London: E. Weaver. Cawdrey, Robert. 2007 [1604]. The First English Dictionary, 1604. Robert Cawdrey’s A Table Alphabeticall. Intro. by J. A. Simpson. Oxford: Bodleian Library. Caxton, William (trans.). 1490. Virgil’s Eneydos. Westminster: William Caxton. Chadwyck-Healey. 2003–2011. Early English Books Online, 1475–1700 (EEBO). Ann Arbor: ProQuest. http://eebo.chadwyck.com/home Cockeram, Henry. 1623. English Dictionarie: or, an Interpreter of Hard English Words. London: Eliot’s Court Press for N. Butter. Cohen, Murray. 1977. Sensible Words: Linguistic Practice in England 1640–1785. Baltimore: Johns Hopkins University Press. Coles, Elisha. 1676. An English Dictionary: Explaining The difficult Terms that are used in Divinity, Husbandry, Physick, Phylosophy, Law, Navigation, Mathematicks, and other Arts and Sciences. London: Samuel Crouch. Wing C 5070. Considine, John. 2008. Dictionaries in Early Modern Europe: Lexicography and the Making of Heritage. Cambridge: Cambridge University Press. Coote, Edmund. 1997 [1596]. The English Schoole-maister. Ian Lancashire, Linda Hutjens, Brent Nelson, Robert Whalen, and Tanya Wood (eds.). Toronto: University of Toronto Library. http://www.library.utoronto.ca/utel/ret/coote/ret2.html (Originally published: London: Widow Orwin for R. Jackson and R. Dexter.) Cotgrave, Randle. 1611. A Dictionarie of the French and English Tongues. London: Adam Islip. Cowell, John. 1607. The Interpreter: or Booke Containing the Signification of Words. Cambridge: John Legate. Crystal, David. 2004. The Stories of English. London: Allen Lane. Elyot, Sir Thomas. 1538. The Dictionary of Syr Thomas Eliot. London: T. Berthelet. Elyot, Sir Thomas. 1542. Bibliotheca Eliotae Eliotis Librarie. London: Thomas Berthelet. Evans, John. 1621. The Palace of Profitable Pleasure. London: W. Stansby. Finkenstaedt, Thomas, Ernst Leisi, and Dieter Wolff (eds.). 1970. A Chronological English Dictionary: Listing 80 000 Words in Order of their Earliest Known Occurrence. Heidelberg: Winter. Finkenstaedt, Thomas, Dieter Wolff, Joachim Neuhaus, and Winfried Herget. 1973. Ordered Profusion: Studies in Dictionaries and the English Lexicon. Heidelberg: Winter. Florio, John. 1598. A Worlde of Wordes, or, Most Copious, and Exact Dictionarie in Italian and English. London: Arnold Hatfield for Edward Blount. Giese, Rachel. 1937. Erasmus’ knowledge and estimate of the vernacular languages. The Romanic Review 28: 3–18. Gill, Alexander. 1619. Logonomia Anglica. London: Iohannes Beale. Go¨rlach, Manfred. 1991. Introduction to Early Modern English. Cambridge: Cambridge University Press. Guy, John. 1988. Tudor England. Oxford: Oxford University Press. Hart, Alfred. 1943. The growth of Shakespeare’s vocabulary. Review of English Studies 19: 242–254. Herttage, S. J. H. (ed.). 1881. Catholicon Anglicum, an English Latin Wordbook, Dated 1483. (Early English Text Society, O. S., 75.) London: Tru¨bner. Huloet, Richard. 1552. Abcedarium Anglico Latinum. Londini: Gulielmi Riddel.
649
650
V Early Modern English Johnson, Samuel. 1979 [1755]. A Dictionary of the English Language. Intro. by Robert W. Burchfield. London: Times Books. Jones, Richard Foster. 1953. The Triumph of the English Language. London: Oxford University Press. Kay, Christian, Jane Roberts, Michael Samuels, and Irene´ Wotherspoon (eds.). 2009. Historical Thesaurus of the Oxford English Dictionary. 2 vols. Oxford: Oxford University Press. Kersey, John. 1702. English Dictionary: Or, a Compleat: Collection Of the Most Proper and Significant Words, Commonly used in the Language. London: Henry Bonwicke and Robert Knaplock. Kyto¨, Merja (ed.). 1996. Manual to the Diachronic Part of the Helsinki Corpus of English Texts: Coding Conventions and Source Texts. 3rd edn. Helsinki: Department of English, University of Helsinki. http://khnt.hit.uib.no/icame/manuals/hc/INDEX.HTM Lancashire, Ian (ed.). 1999. The Early Modern English Dictionaries Database (EMEDD). http:// homes.chass.utoronto.ca/~ian/emedd.html Lancashire, Ian. 2002. “Dumb Significants” and Early Modern English definition. In: Jens Brockmeier, Min Wang, and David R. Olson, Literacy, Narrative and Culture, 131–154. Richmond, Surrey: Curzon. Lancashire, Ian. 2006. Law and Early Modern English lexicons. In: R. W. McConchie, Olga Timofeeva, Heli Tissari, and Tanja Sa¨ily (eds.), HEL-LEX: New Approaches in English Historical Lexis, 8–23. Somerville, MA: Cascadilla Press. http://www.lingref.com/cpp/hel-lex/2005/ paper1342.pdf Lancashire, Ian (ed.). 2006–. Lexicons of Early Modern English (LEME). Toronto: University of Toronto Library and University of Toronto Press. http://leme.library.utoronto.ca/ Levins, Peter. 1867. Manipulus Vocabulorum: A Rhyming Dictionary of the English Language, by Peter Levins. (1570). H. B. Wheatley (ed.). (Early English Text Society, O. S., 27.) London: Tru¨bner. Liberman, Anatoly. 1998. An annotated survey of English etymological dictionaries and glossaries. Dictionaries 20: 21–96. Lily, William and John Colet. 1549. A Short Introduction of Grammar, 1549. (English Linguistics 1500–1800, ed. by R. C. Alston, 262.) Menston: Scolar Press, 1970. Locke, John. 1690. An Essay Concerning Humane Understanding. London: for Thomas Basset by Edward Mory. Matthew, Colin, Brian Harrison, and Lawrence Goldman (eds.). 2004–08. Oxford Dictionary of National Biography Online. Oxford: Oxford University Press. http://www.oxforddnb.com/ index.jsp Mayhew, Anthony Lawson (ed.). 1908. The Promptorium Parvulorum. The First English Dictionary. c. 1440 A. D. (Early English Text Society, E.S., 102.) London: Kegan Paul, Trench, Tru¨bner, and Henry Frowde, Oxford University Press. McConchie, R. W. 1997. Lexicography and Physicke: The Record of Sixteenth-century English Medical Terminology. Oxford: Clarendon Press. McDermott, Anne. 2002. Early dictionaries of English and historical corpora: In search of hard words. In: Javier A. Dı´az Vera (ed.), A Changing World of Words: Studies in English Historical Lexicography, Lexicology and Semantics, 197–226. Amsterdam/Atlanta, GA: Rodopi. Minsheu, John. 1617. Ductor in Linguas. London: John Browne. Mulcaster, Richard. 1582. The First Part of the Elementarie. London: Thomas Vautroullier. Nevalainen, Terttu. 1999. Early Modern English lexis and semantics. In: Roger Lass (ed.), The Cambridge History of the English Language. Vol. III. 1476–1776, 332–458. Cambridge: Cambridge University Press. Nevalainen, Terttu. 2006. An Introduction to Early Modern English. Edinburgh: Edinburgh University Press. Nevalainen, Terttu, Helena Raumolin-Brunberg, Jukka Kera¨nen, Minna Nevala, Arja Nurmi, and Minna Palander-Collin. 1998. Corpus of Early English Correspondence (CEEC). Department
40 Early Modern English: Lexicon and semantics of English, University of Helsinki. http://www.helsinki.fi/varieng/CoRD/corpora/CEEC/index. html Nowell, Laurence. 1952. Laurence Nowell’s Vocabularium Saxonicum. Albert H. Marckwardt (ed.). Ann Arbor: University of Michigan Press. Nurmi, Arja. 1998. Manual for the Corpus of Early English Correspondence Sampler. Helsinki: Department of English, University of Helsinki. http://kh.aksis.uib.no/icame/manuals/ceecs/ INDEX.HTM Palsgrave, John. 2003 [1530]. L’e´claircissement de la langue franc¸aise, 1530: texte anglais original / John Palsgrave. Susan Baddeley (ed. and trans.). Paris: Champion. Phillips, Edward. 1658. The New World of English Words, 1658. (English Linguistics 1500–1800, ed. by R. C. Alston, 321.) Menston: Scolar Press, 1969. Pollard, A. W. and G. R. Redgrave. 1976–91. A Short Title Catalogue of Books Printed in England, Scotland, & Ireland, and of English Books Printed Abroad, 1475–1640. 2nd edn. 3 vols. William A. Jackson, F. S. Ferguson, and Katherine F. Pantzer (eds.) London: The Bibliographical Society. Ray, John. 1674. A Collection of English Words. Not Generally used, with their Significations and Original. London: H. Bruges. Rider, John. 1589. Bibliotheca Scholastica. Oxford: Joseph Barnes. Ridley, Mark (attributed). 1996. A Dictionarie of the Vulgar Russe Tongue Attributed to Mark Ridley, Gerald Stone (ed.). Ko¨ln/Weimar/Wien: Bo¨hlau Verlag. Rissanen, Matti, Merja Kyto¨, Leena Kahlas-Tarkka, Matti Kilpio¨, Saara Nevanlinna, Irma Taavitsainen, Terttu Nevalainen, and Helena Raumolin-Brunberg. 1991. The Helsinki Corpus of English Texts. In: ICAME Collection of English Language Corpora (CD-ROM), 2nd edn., Knut Hofland, Anne Lindebjerg, and Jørn Thunestvedt (eds.), The HIT Centre, University of Bergen, Norway. For manual, see http://khnt.hit.uib.no/icame/manuals/hc/index.htm Rissanen, Matti, Merja Kyto¨, and Minna Palander-Collin (eds.). 1993. Early English in the Computer Age: Explorations through the Helsinki Corpus. Berlin/New York: Mouton de Gruyter. Salmon, Vivian. 1990. Some views on meaning in sixteenth-century England. In: Peter Schmitter (ed.), Essays towards a History of Semantics, 33–53. Mu¨nster: Nodus. Scha¨fer, Ju¨rgen. 1980. Documentation in the OED: Shakespeare and Nashe as Test Cases. Oxford: Clarendon Press. Scha¨fer, Ju¨rgen. 1989. Early Modern English Lexicography. 2 vols. Oxford: Clarendon Press. Shakespeare, William. 1997. The Riverside Shakespeare. 2nd edn. Boston: Houghton Mifflin. Sheppard, Francis. 1998. London: A History. Oxford: Oxford University Press. Sidney, Sir Philip. 1595. The Defence of Poesie. London: William Ponsonby. Simpson, John (ed.). 2000–. The Oxford English Dictionary. 3rd edn. online. Oxford University Press. www.oed.com/ Skinner, Thomas. 1671. Etymologicon Linguæ Anglicanæ. London: T. Roycroft. Slaughter, Mary M. 1982. Universal Languages and Scientific Taxonomy in the Seventeenth Century. Cambridge: Cambridge University Press. Somner, William. 1659. Dictionarium Saxonico-Latino-Anglicum. Oxford: William Hall for Daniel White. Spevack, Marvin. 1973. The Harvard Concordance to Shakespeare. Cambridge, MA: Belknap Press of Harvard University Press. Stein, Gabriele. 1985. The English Dictionary before Cawdrey. Tu¨bingen: Niemeyer. Stein, Gabriele. 1986. Definitions and first-person pronoun involvement in Thomas Elyot’s dictionary. In: Dieter Kastovsky and Aleksander Szwedek (eds.), Linguistics across Historical and Geographical Boundaries, 1465–1474. Berlin/New York: Mouton de Gruyter. Thomas, Keith. 1986. The meaning of literacy in Early Modern England. In: Gerd Baumann (ed.), The Written Word: Literacy in Transition, 97–131. Oxford: Clarendon Press. Thomas, R. J., Gareth A. Beven, and P. J. Donovan. 1967–2002. Geiriadur Prifysgol Cymru. A Dictionary of the Welsh Language. 4 vols. Cardiff: University of Wales Press.
651
652
V Early Modern English Thomas, Thomas. 1587. Dictionarium Linguae Latinae et Anglicanae. Cambridge: Richard Boyle. Turner, William. 1999. Libellus de Re Herbaria Novus. Mats Ryde´n, Hans Helander, and Kerstin Olsson (eds. and trans.). Uppsala: Almqvist & Wiksells. Waite, Greg (ed.). 2008–. The Textbase of Early Tudor English. Dunedin, NZ: University of Otago. http://www.hlm.co.nz/tudortexts/. Waswo, Richard. 1987. Language and Meaning in the Renaissance. Princeton: Princeton University Press. Welzenbach, Rebecca A. 1999–. Early English Books Online Text Creation Partnership (EEBOTCP). http://www.lib.umich.edu/tcp/eebo/description.ht. Williams, Franklin B., Jr. 1962. Index of Dedications and Commendatory Verses in English Books before 1641. London: Bibliographical Society. Wilson, Thomas. 1553. The Arte of Rhetorique. London: Richardus Grafton. Wing, Donald. 1972–88. Short-Title Catalogue of Books Printed in England, Scotland, Ireland, Wales, and British America, and of English Books Printed in Other Countries, 1641–1700. 2nd edn. 3 vols. New York: The Index Committee of the Modern Language Association of America. Withals, John. 1553. A Shorte Dictionarie for Yonge Begynners. London: T. Berthelet. Worlidge, John. 1669. Dictionarium Rusticum. London: T. Johnson.
Ian Lancashire, Toronto (Canada)
41 Early Modern English: Pragmatics and discourse 1 2 3 4 5 6
Introduction Speech acts in the EModE period Im/politeness studies relating to the EModE period Discourses and discourse strategies Summary References
Abstract This chapter provides an overview of the most widely studied discoursal/pragmatic phenomena of Early Modern English discourse: speech acts, address terms, politeness, discourse markers, and (other) discourse strategies. The Early Modern English period is assumed here to cover approximately three hundred years, starting shortly after the introduction of the printing press to England in 1476; thus, some scholars may want to place several of the studies included here within the late Early Modern English period or at the beginning of the Modern English period (see, e.g., Jucker 2000: 7; Go¨rlach 1991: 9; see also Lewis, Chapter 56). This chapter also highlights (researchers’ awareness of) the importance of considering the “interplay between the individual, language and society” (Palander-Collin 2010) when seeking to explain discoursal/pragmatic phenomena of times past.
1 Introduction There is a lack of consensus in respect to what constitutes Historical Pragmatics, Historical Discourse Analysis, and Historical Dialogue Analysis as disciplines, Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 652–667
41 Early Modern English: Pragmatics and discourse and the extent to which they overlap (for useful discussions of their similarities and differences see Brinton 2001; Jacobs and Jucker 1995; Jucker et al. [eds.] 1999; Traugott 2004). One point of similarity relevant to this chapter, however, is the attention given (by advocates of the three disciplines) to the phenomena of speech acts, address terms, politeness, discourse markers, and discourse strategies in the Early Modern English (henceforth EModE) period. Indeed, the aforementioned constitute the most widely studied discoursal/pragmatic phenomena (within the EModE period) and, as such, are to be the focus of this chapter, beginning with an investigation of speech acts (see Section 2). Where relevant, I will also discuss the (prevalent) use of corpus-linguistic techniques within the various fields.
2 Speech acts in the EModE period The study of speech acts in their historical context is older than one might imagine if we include non-English studies. For example, Schlieben-Lange and Weydt began investigating the historicity of Romance speech acts in the 1970s (see, e.g., Schlieben-Lange 1976, 1983; Schlieben-Lange and Weydt 1979). The bulk of speech act studies relating to the EModE period date from the 1990s onwards, however, and focus on: – directives such as requests and orders – that is, verbal acts whereby S (= Speaker) attempts to get H (= Hearer) to do something (e.g. Rudanko 1993; Busse 2002, 2008; Culpeper and Archer 2008); – expressives such as insults, curses, compliments, and thanks – that is, verbal acts whereby S expresses a psychological state towards H or in respect to a proposition (e.g. Jacobsson 2002; Culpeper and Semino 2000; Jucker and Taavitsainen 2000; Taavitsainen and Jucker 2008b); – commissives such as promises – that is, verbal acts whereby S commits to do some future act(s) for H (e.g. Arnovick 1994, 1999; Valkonen 2008); – “rogatives” (i.e. questions) – that is, verbal acts whereby S seeks information from H (e.g. Claridge 2005; Archer 2005). The above studies draw on literary sources as well as sources that are taken to be representative of “real-life” speech events, as the former are considered to be a valuable means by which to investigate past speech events (for further discussion, see Jucker et al. [eds.] 1999: 16; Jucker and Taavitsainen 2003: 8–9). For example, Busse (2008) provides an inventory of directives in Shakespeare’s King Lear; Rudanko (1993) investigates questions and requests in Shakespeare’s Othello and Coriolanus; Culpeper and Archer (2008) explore requests in comedy plays and courtroom proceeding transcripts taken from the Sociopragmatic Corpus (1640–1760); Archer (2005) draws from the courtroom proceedings of the Sociopragmatic Corpus to study questions; Claridge (2005) explores questions, using the Lampeter Corpus of Early Modern English Tracts (1640–1740); and Jacobsson (2002) identifies expressions associated with thanking from the six text-types which make up the Corpus of English Dialogues 1560–1760 (plays, fiction, didactic works, language teaching texts, witness depositions, and courtroom proceedings).
653
654
V Early Modern English
2.1 Methodologies and approaches utilized Schlieben-Lange’s (1976, 1983) work on Romance speech acts enabled her to pioneer an approach that has been hugely influential for English speech act studies (see, e.g., Arnovick’s [1999] culturally-rich descriptions of insults, promises, curses and partings). Schlieben-Lange’s (form-based) approach involves attending to both “diachronic process” and “historical contrast”: that is, gleaning evidence of verbal activity from texts through the identification of performative speech-act verbs, and then confirming the historical appropriacy of the (speech act) labels utilized via historical dictionaries and additional contextual information/evidence (“histories of institutions”, law treatises, etc.). More recently, Kohnen (2002: 238, 241) has made a distinction between formbased approaches that engage in “structured eclecticism” and form-based approaches that engage in “illustrative eclecticism”. The latter involves researchers picking out relevant examples of a given speech act in the respective periods covered by their particular study on the basis of their (intuitive) reading of the data and of the relevant literature (see, e.g., Jucker and Taavitsainen 2000). In contrast, structured eclecticism involves “a [more] deliberate [and systematic] selection of typical patterns which [researchers] trace by way of a representative analysis throughout the history of English” (Kohnen 2002: 238). As one might expect, given the emphasis on “typical patterns”, structured eclecticism (i) tends to be adopted when studying speech acts that occur in routinized forms, and (ii) is the suggested/favored approach when engaging in computerized searches for specific speech acts (Kohnen 2002). Indeed, Jucker and Taavitsainen (2008) have shown how, because Present-day English realizations of apologies are highly routinized and depend on a small range of conventionalized lexical forms, these patterns can offer an excellent point of departure from which to (seek to) identify apologies in Renaissance data. Similarly, Valkonen (2008) has sought to identify promises in the ARCHER corpus and the Chadwyck-Healey Eighteenth Century Fiction database, using a small set of lexical-morphosyntactic search patterns (such as PROMIS*, PLEDG*, VOW*, SWEAR*, SWORN, VOUCH*, GAURANTE*), which were originally derived from Wierzbicka’s (1987) list of (verbs for) promises. As an approach based on structured eclecticism will only allow for the automatic detection of performative instances of speech acts, and Culpeper and Archer (2008) wanted to capture all instances of directives (direct and indirect) in their Sociopragmatic Corpus, they have developed a (manually applied) annotation scheme as a means of identifying directives in their drama and trial data. Although directives are regarded as being less sensitive to cultural/historical variation than speech acts such as apologies, complaints and compliments (Kohnen 2002), Culpeper and Archer (2008: 58, 47–48) nevertheless advocate a “multi-feature view of speech acts”, which makes use of: – formal features (such as particular conventionalized pragmalinguistic strategies or Illocutionary Force Indicating Devices [IFIDs]: see, e.g., Searle 1975; Aijmer 1996); – contextual beliefs (such as Searle’s [1969] observation that it is not obvious that [the desired] future action will be performed by the target in the normal course of events); – co-textual features (such as pre-requests: see, e.g., Edmonson and House 1981; Levinson 1983: 356–364; Tsui 1994: 110–111); – outcomes (such as the target performing the action specified in the earlier speech act: see, e.g., Austin 1962).
41 Early Modern English: Pragmatics and discourse This effectively equates to capturing speech-act function and context-of-utterance as well as form and, as such, seeks to avoid “pragmatic false friends”, i.e., constructions such as Ford’s question to Falstaff (Would you speak with me?), “which, against a contemporary background, [can] suggest a wrong pragmatic interpretation”; for example, a request for Falstaff ’s attention as opposed to a clarification question along the lines of Did you want to talk to me? (see Kohnen 2002: 239–240).
2.2 Directives in the EModE period: assessing in/directness Culpeper and Archer (2008) have also sought to determine the diachronic applicability of the in/directness approach first advocated by Blum-Kulka et al. (eds.) (1989), as Blum-Kulka and House (1989) have suggested that conventional indirectness may be “universal” in our modern world (in/directness, as utilized here, does not relate to Searle’s [1975] non-mapping of form and function, but to the transparency with which the illocutionary point of a particular utterance is signalled). Culpeper and Archer’s (2008) findings appear to invalidate Blum-Kulka and House’s (1989) claim, however, for more than two-thirds of the directives in their late EModE data were achieved directly, via impositive strategies, the majority of which utilized an imperative form (such as Take away her sword). Moreover, when conventionally indirect strategies were in evidence, they were used by relatively powerful people/intimates of high-status, rather than by people with less power, as we might expect today. These “powerful” interactants also exhibited a perceived preference for directives marked by contextual beliefs – specifically, volition (cf. I would speak with you) or obligation (cf. We must go to the city) – a factor that contrasts starkly with a key feature of present-day conventionally-indirect directives: modern-day (British) users are said to orient their requests to the preparatory condition of ability (i.e. a contextual belief, relating to the ability of the target to perform a future action: cf. the widely-discussed request, Can you pass the salt?). Culpeper and Archer (2008) point to the “democratisation of discourse” (see Fairclough 1993: 98) and the subsequent reduction in overt markers of power asymmetry between people of unequal institutional power as a possible explanation for the change(s). A perceived need in the EModE period to signal (a level of) volition and/or obligation in one’s directives may also help to explain the prevalence of let-requests in the Sociopragmatic Corpus, in particular, and the EModE period more generally (see Kohnen 2004), as let requests exhibited both volition and obligation, depending on whether they requested that something be “allowed” or suggested that something be “desired”.
2.3 The “fuzziness” of speech acts/speech act categories Jucker and Taavitsainen (2000) have recently suggested a “fuzziness” approach to speech act research, which allows for both synchronic and diachronic variation. What this means, in practice, is a rejection of the application of speech act theory (in its most conservative Searlean sense) in favor of a view of speech acts as prototypes linked by a shared “pragmatic space”. For example, they argue that insults (synchronic and diachronic) share the same starting point: they contain[ed] “a predication about the target” that is/was “disparaging” in some way (Taavitsainen and Jucker 2008a: 6). However, (the actual realizations of) insults are/were located across several
655
656
V Early Modern English clines, for example, form (“ritual/rule governed-creative”), context-dependence (“conventional-particular”), and speaker attitude (“ludic-aggressive”, “intentionalunintentional”), etc. They go on to suggest that the framework they initially drafted for insults can also be applied to other speech acts/speech act categories as long as we allow the latter to dictate which of the components of the multidimensional framework are foregrounded and which are backgrounded: for example, Taavitsainen and Jucker predict that, with apologies, the form category is likely to be most prominent, as apologies tend to be “expressed in routinised, perhaps even ritual and rule-governed forms”; the dimension of “irony versus sincerity” (a third aspect of speaker attitude) will become “prominent with compliments”; and speaker attitude is also likely to be “of special concern” when constructing directives (Taavitsainen and Jucker 2008a: 7; cf. Culpeper and Archer 2008). The benefit of a “fuzziness” approach is that it allows for the fact that individual instances of a particular speech act may vary in their degree of conformity to the prototypical manifestation of that act – to the extent that their group identity can be vague. Archer (2005: 127–128, 339–340) therefore opts to identify macro speech act categories when documenting the speech acts that were utilized in the historical English courtroom (1640–1760). By way of illustration, commissive/directive speech acts are captured by the “Counsel” macro speech act category, and include caution, warn, threaten, coerce, advise and recommend; directive (and, more specifically, requirement) speech acts are captured by the “Require” macro speech act category, and include command, order, direct and demand; and directive (and, more specifically, requestive) speech acts are captured by the “Request” macro speech act category, and include desire, plead, beseech, implore and appeal.
3 Im/politeness studies relating to the EModE period The speech acts that have received most attention to date (in both a synchronic and diachronic context) are those that constitute face-threatening acts (henceforth FTAs); requests, apologies, complaints, and thanks. For example, Busse (2008), Jacobsson (2002) and Taavitsainen and Jucker (2008b) are amongst a number of speech act researchers to make use of Brown and Levinson’s (1987 [1978]) politeness theory to explain EModE requests, thanks, and compliments. As Brown and Levinson’s theory promotes Goffman’s (1967: 5) idea that every rational human being understands s/he and his/ her conversational interactant(s) have face wants that are maintained/enhanced using a variety of (super-)strategies, it has also proved useful to Fitzmaurice (2000), when demonstrating how the modal verbs can, may and will were manipulated during the EModE period so as to express both tentativeness and insistence (see Section 3.2.). The majority of researchers, however, (e.g. Brown and Gilman 1989; Calvo 1992; Hope 1994; Kopytko 1993, 1995; Nevalainen and Raumolin-Brunberg 1995; Raumolin-Brunberg 1996; Nevala 2004; Walker 2007) have utilized Brown and Levinson’s approach to investigate EModE address terms (including pronominal usage). Brown and Gilman’s (1989) exploration of Shakespeare’s use of address terms in Hamlet, King Lear, Macbeth, and Othello is probably the most well-known politeness study of the EModE period. Studies involving you and thou include Hope’s (1994) investigation of (conversations recounted in) depositions made to the Durham ecclesiastical court in the 1560s, and Walker’s (2007) investigation of trial, drama, and witness deposition data
41 Early Modern English: Pragmatics and discourse (taken from the Corpus of English Dialogues). Walker’s work, in particular, documents the relationship between the demise of thou during the EModE period and factors such as the context-of-utterance and the age, sex and rank of the interlocutors (cf. Brown and Gilman 1960: 225–227).
3.1 Address forms and politeness: a case study of the Bacons One of the main findings in respect to EModE address formulae is that – like their modern counterparts – they exhibit positive and negative politeness (see, e.g., Nevalainen and Raumolin-Brunberg 1995; Raumolin-Brunberg 1996; Nevala 2004). For example, Nevala (2004) investigated the (early) 17th-century correspondence of a prestigious Norfolk family (the Bacons) as a means of determining the way(s) in which relative power and social distance affected the forms of address/reference utilized in respect to Sir Nicholas (Lord Keeper for Elizabeth I and head of the Bacon family) and his eldest son, Nathaniel (cf. Brown and Levinson’s [1987 (1978)] argument that power, distance and the FTA’s “ranking of imposition” determine politeness levels). Nevala (2004: 2138) found that, when directly addressing Sir Nicholas, writers utilized strategies denoting “extreme negative politeness, regardless of the distance” between them, and concluded that this was due to Sir Nicholas’s superior social status, as well as his role as head of the family (in what was a highly stratified society). In contrast, the direct address that Nathaniel Bacon received from his correspondents was very much dependent on (his relationship with) them: nuclear family members tended to use positive politeness strategies (i.e. first name, last name, kinship term or a combination of the three); family servants utilized negative politeness (i.e. right worshipful, your worship, or your honour); (more) distant writers alternated between neutral and negatively polite formula (i.e. cousin Bacon, Sir, Right Worshipful ). Nevala (2004: 2147) contends that Brown and Levinson’s model therefore accounts well for the usage of EModE address terms. But she is less convinced of its value in respect to reference terms: nuclear family members would occasionally use both kinship term and title to identify Sir Nicholas, for example. Nevala also found evidence of non-kin writers (both equal in power and subservient to the referent) using kinship terms, which leads her to suggest that the choice of referential term(s) might have been more conscious than the choice of address term(s). Nevala (2004: 2149) further suggests that the possibility of letters being read by others (and thus the need to accommodate to the terms likely to be used by not only the addressee but also those persons who are not overtly referred to in the letters) may account for this. She goes on to propose a further axis for describing EModE referents (addressee-referent), to complement the axes previously identified by Brown and Levinson (1987 [1978]) and Comrie (1976) in respect to modern data (the speaker-addressee axis, the speaker-referent axis, the speaker-bystander axis and the speaker-setting axis).
3.2 Is “deference” the same as “politeness” in an EModE context? Brown and Levinson (1987 [1978]) believe that, in a modern context, giving deference links to power and, as such, provides a further example of negative politeness (see also Leech’s 1983 approbation and modesty maxims). Watts (2003: 80, 176) disagrees, contending that deference is not politeness if it constitutes the norm for a particular activity
657
658
V Early Modern English type: rather, it equates to “politic” behavior. Nevala (2004) is not convinced of the merit of the “politic” approach to deference in a diachronic context, however, as she believes that address formulae could (and did) “work within, and for, politeness” in the EModE period (see also Raumolin-Brunberg [1996: 168], who argues that the choice of address form was “never […] predictable” in the Corpus of Early English Correspondence). Klein’s (1994a, 1994b) studies of the development of linguistic thinking as it related to politeness in the late 17th and early 18th centuries may offer support to Nevala and Raumolin-Brunberg’s stance – especially given the affinity of some of Klein’s findings with Brown and Levinson’s (1987 [1987]) idea of the “model person”: Klein emphasizes the link between a growing politeness awareness and an ideology of standardization, which (rather than merely promoting lexical or grammatical correctness) sought to address “the question of effective language use […] founded on the conditions of gentlemanly and urban conversation” (Klein 1994a: 43). The Third Earl of Shaftesbury’s promotion of a “model of gentlemanly conversation” (Klein 1994a: 43) is thought to have been particularly influential. Klein (1994b) also highlights the importance of the socio-historical conditions, in particular: a. the new discursive/socializing conditions afforded by the declining influence of Church and Court, b. the newly-emerging patterns of urban development (which would house not only the aristocrats and greater/lesser gentry, but also the “pseudo-gentry, professionals and commercial elements” [Klein 1994b: 11]), and c. the growth of print media/the deregulation of the press (after 1695). The cumulative effect, according to Klein (1994b: 12), was that politeness became “a model of cultural action” that not only shaped “talk” in its literal sense, but also helped to “shape a wide range of cultural institutions and practices in the eighteenth century”, from letter writing (see, e.g., Fitzmaurice 2000) to educational and scientific treatises (see, e.g., Atkinson 1999). For example, Fitzmaurice (2000) draws on Boyer’s (1702: 106, 108) idea of politeness as “a dextrous management of our Words and Actions, whereby we make other People have better Opinions of us and themselves” (cited in Klein 1994b: 4), as a means of justifying Margaret Cavendish’s apparent manipulation of the (multiple) semantic-pragmatic meanings of the modals can, may and will in her CCXI Social Letters. Cavendish’s purpose, it seemed, was to construct a voice (for herself and/or her writing persona) that was not “completely authoritative [n]or wholly tentative” and, as such, was attentive to the face issues of herself/her persona(e) and her audience (Fitzmaurice 2000: 8). Put simply, the “art” of letter writing, at this time, appeared to involve balancing the notions of “appropriate discourse and proper stance” in a way that allowed for “the spirit of philosophical enquiry” (Fitzmaurice 2000: 17–18). As Watts (2003: 40) highlights, the main focus of Klein’s (1994a, 1994b) work is politeness1 as opposed to politeness2: that is, it captures the way(s) in which polite behavior was evaluated and commented on by lay members of the EModE language community – and, more particularly, the “gentrification” that politeness underwent at this time (in order to appropriate a new hegemonic discourse). I would therefore contend that we need more studies that seek (like Fitzmaurice 2000) to explain the linguistic consequence(s) (i.e. the politeness1) of the so-called “gentrification of
41 Early Modern English: Pragmatics and discourse politeness”, if we are to fully appreciate EModE conceptions of facework (i.e. the communicative strategies that interactants used to preserve and/or contravene socially-appropriate behavior).
3.3 EModE impoliteness studies Recently, researchers have begun to investigate instances of face attack (as opposed to face maintenance), using different Shakespeare plays (see, e.g., Culpeper 1998; Rudanko 2006; Bousfield 2007). For example, Bousfield explores a conversation between two of the main characters of Henry IV, Part I, where Hal and Falstaff take on the identities of the characters of Hal’s father (King Henry IV) and Hal respectively (II.iv). Using Leech’s (1983: 142–144) definition of banter as his starting point (i.e. saying something which is obviously untrue and obviously impolite as a means of achieving solidarity/social bonding with one’s interlocutor), Bousfield shows how Shakespeare intersperses the dialogue of Hal-as-King with not only banter – that is, comments about Falstaff that are obviously untrue – but also comments that are actually true and, moreover, extremely offensive. Thus, Falstaff (who is known to be diseased) is described by Hal-as-King as being “a trunk of humours” and a “swollen parcel of dropsies”. What we are describing here is a stylistic device utilized by Shakespeare to aid characterization. Nevertheless, Bousfield (2007: 216) is careful to remind readers of Mills’s (2003: 14) observation that we can manipulate banter in a modern (and, it would seem, a historical) context so that, although what we say is superficially non-serious, it is actually closer to our “true feelings than perceived to be by H”. As such, it becomes “surreptitiously face damaging”. Kryk-Kastovsky (2006) and Archer (2008) draw from EModE courtroom transcripts for their discussions of face attack. Archer does so with an important caveat: that existing impoliteness models (e.g., Culpeper 1996, 2005; Culpeper et al. 2003; Bousfield 2008) be extended as a means of capturing verbal aggression more generally, and not just intentional impoliteness. Her proposal involves utilizing Goffman’s (1967: 14) “intentional”, “incidental”, and “unintended” categories relating to face threat, as opposed to the first only, as Culpeper et al. (2003: 1550) do (as a means of identifying impoliteness). She argues that this would allow us to differentiate between malicious face attacks (where the primary intent was to cause damage), and instances when defendants and lawyers undertook verbal acts which they knew might have “offensive consequences” (cf. Goffman 1967: 14).
4 Discourses and discourse strategies Modals, address formulae, and implicature were not the only means by which face was managed in the EModE period, according to Culpeper and Kyto¨ (1999): hedges such as about and discourse markers (henceforth DMs) such as well and why were also used to strategically manage face (as well as information, discourse, and style). For Culpeper and Kyto¨ (1999: 299), style is “involved whenever a hedge is used”, regardless of the activity/text-type: it relates to the interactants’ understanding of the in/formality of their particular (discoursal) situation, putting an interactant at ease and/or creating a sense of involvement/solidarity, etc. (all of which also appear to relate to Brown and Levinson’s [1987 (1978)] concept of “positive politeness”). In addition, Culpeper and
659
660
V Early Modern English Kyto¨ (1999: 304–305) identify an interesting relationship between activity (or text) types and specific EModE hedges: for example, about “occurred with striking frequency” in the EModE witness depositions within the Corpus of English Dialogues; this is presumably because it offered the user a strategic means of adding “fuzziness to a claim” without appearing to be deliberately withholding information (cf. “[…] about a month agoe, as this Examinate was coming towards his Mothers house […] about ten Roodes distant from the same house: and about two or three nights after […]”: taken from a deposition relating to the 1612 examinations of the Pendle Witches). Similarly, the DMs why and well occurred very frequently in their drama data, and appeared to “play a role in managing that discourse” (Culpeper and Kyto¨ 1999: 305): why, for example, was used to signal (i) S’s surprise and/or “challenge something the previous speaker had said” and, (ii) a change of speaker (see Section 4.1).
4.1 Discourse Markers Culpeper and Kyto¨’s identification of primarily pragmatic/discoursal strategies is not surprising, given that hedges “help signal speakers’ feelings and attitudes to their messages, their co-participants and the situation as a whole” (Nikula 1996: 11–12; cited in Culpeper and Kyto¨ 1999: 294). Brinton confirms that some types of DMs share the pragmatic (textual/expressive), procedural (inferential), metalinguistic, and/or sociolinguistic functions of hedges (e.g. relate to turn-taking and politeness), and also lists their phonological, syntactic and semantic features (Brinton 1996): – DMs are phonologically short items that normally occur sentence-initially (but can occur in medial or final position; see, e.g., Fraser 1999: 938); – DMs are syntactically independent elements (often constituting separate intonation units) which occur with high frequency (in oral discourse in particular); – DMs lack semantic content (i.e. they are non-referential/non-propositional in meaning); – DMs are non-truth-conditional elements, and thus optional (i.e. they may be deleted). Fraser (1999: 946, 950) further suggests that (modern-day) DMs tend to be “drawn from the syntactic classes of conjunctions, adverbials, or prepositional phrases”. As such, although DMs constitute a pragmatic class (in the sense that their more specific interpretation is “negotiated” by the context), they nevertheless maintain “the syntactic properties associated with their class membership”: but, for example, will denote its core meaning of “simple contrast” plus whatever additional interpretation is warranted, given the context-of-use.
4.2 Identifying DMs of times past Historically-focused research relating to DMs can be synchronic (i.e. researchers explore a particular time-period that is fixed (be it years, decades, centuries, etc.) or diachronic (i.e. they explore successive synchronic stages of time). Whether synchronic or diachronic, most studies focus not only on “what they [DMs] are, what they mean, and what function(s) they manifest” (Fraser 1999: 933) but also on how individual DMs
41 Early Modern English: Pragmatics and discourse “pattern” and how those patterns have been derived over time. For example, Brinton (2006) has identified the following syntactic pathways as ones which might lead to DM development: – – – –
adverb/preposition > conjunction > DM; predicate adverb > sentential adverb structure > parenthetical DM (cf. Traugott 1995); imperative matrix clause > indeterminate structure > parenthetical DM; relative/adverbial clause > parenthetical DM.
In contrast, Traugott and Dasher (2002) have identified pathways that are more semantic in orientation, namely: – truth-conditional > non-truth conditional > non-truth conditional, – content > procedural, and nonsubjective > subjective > intersubjective meaning, – scope within the proposition > scope over the proposition > scope over discourse. Diachronic studies of DMs capturing the EModE period include Jucker’s (1997) study of well, Fischer’s (1998) study of marry, and Schwenter and Traugott’s (2000) study of in fact. For example, Jucker (1997: 93) traces the development of well from the Old English to the Modern English period and concludes that it exhibits both a textual and an intertextual function historically, but that “one function typically predominates over the other” (cf. Traugott [1982], who shows how well (right and why) move from a propositional through a textual to an interpersonal meaning). Marry also exhibited textual and intertextual functions, according to Fischer (1998); the former, at the beginning of a turn, and the latter, as a means of expressing a range of speaker attitudes. Drawing on the OED, Fischer (1998) initially shows how marry is first attested in the second half of the 14th century, as an “oath or an ejaculatory invocation” (to the Virgin Mary). He goes on to identify a growth in the use of marry during the period 1570– 1640, but it had come to function as an interjection as opposed to a religious invocation. Fischer suggests that these tendencies (i.e. developing into a pragmatically weaker interjection/losing any religious association) may also account for the (historical development of) other religious oaths. Schwenter and Traugott (2000) adopt a similar formto-meaning approach to show how the DM in fact came to take on epistemic modal meanings in contrastive (or “dialogic”) contexts (that is, contexts where the speakers/ writers (SPs/Ws) presented opposing arguments to real or imaginary interlocutors). The development of in fact is shown to be as follows by Schwenter and Traugott (2000: 12): manner adverb > adversative adverb (expressing contrast to someone else’s or to the locutor’s own prior proposition) > elaborating DM (signalling that what follows is a stronger argument than what precedes, with respect to SP/W’s purpose at the point in the discourse). We have also come to associate the all I said/did was X pattern with adversitivity and refutation: the specificational all-cleft construction (with an ALL-NP-V-BE-X string) arose around 1600 (in dialogic contexts), and then quickly moved from an ‘everything’ meaning to an ‘only’ meaning with a clausal focus, becoming conventionalized in the later part of the 17th century (i.e. the ‘only’ meaning became semanticized into the construction). As such, the construction exhibited a similar switchreferencing characteristic to modern all-clefts (cf. Bonnelli’s 1992 “change of posture”). That is to say, “in most cases the most important function appears not to be to fill an open
661
662
V Early Modern English proposition, but rather to highlight an upcoming statement as salient, and impose an exhaustive reading on it” (Traugott 2008: 162). Traugott (2008: 159) provides us with the following example (1), dating from 1658, taken from the Chadwyck Healey website (http://lion.chadwyck.com, last accessed 12 January 2010): (1)
[…] he doth me notorious wrong, I did not mention any Principles of Vnity in this place […] All I said was this, That we doe not separate from other Churches, but from their Accidentall Errours (1658 Bramhall, Schisme Garded)
EModE synchronic studies of DMs include Taavitsainen (1995), Blake (1992, 1996), Brown and Gilman (1989), Kryk-Kastovsky (1998), and Busse (1999). Using representations of direct speech taken from literature texts, Taavitsainen (1995: 439, 465) shows how some interjections (such as ah, alas, benedicte, eh, fie, and tush) exhibited DM-like characteristics: that is, “they encode[d] speaker attitudes and communicative intentions”, they sought to deliberately manipulate “reader involvement”, and they also served textual functions (but see Fraser [1999: 943], who argues that interjections are best regarded as pragmatic idioms). Blake (1992: 1996) investigates the discourse markers why and what in Shakespeare’s Henry VI Part 3, Two Gentlemen of Verona, and Othello, showing how the former was used to draw a logical conclusion from what has gone before (and often gave a tone of superiority and potential disparagement), and the latter was used to express surprise or incredulity (and, on occasion, to express contempt or scorn). Kryk-Kastovsky (1998), Brown and Gilman (1989) and Busse (1999) have all analysed the DMs pray and prithee; Kryk-Kastovsky, in transcripts of the EModE trials of Titus Oates and Lady Alice Lisle, and Brown and Gilman and Busse, in Shakespearean plays. Kryk-Kastovsky found that pray and prithee were amongst the most frequent of the DMs in her trial data (although they were not overly frequent). Pray and prithee constituted shortened forms of their fuller (propositionallyintact) forms I pray you and I pray thee in the EModE period; their purpose was to add deference to, for example, questions and requests. This makes them ideally suited to the (historical) courtroom context. That said, Brown and Gilman and Busse have also found that they were used quite extensively by Shakespeare. They disagree, however, on whether they should be seen as in-group identity markers or deference markers (for further detail of these and other EModE DMs, see Jucker 2002; for diachronic studies, see Akimoto 2000; Traugott and Dasher 2002: 252–255; Brinton 2010).
5 Summary This chapter has documented studies that have utilized dramatic dialogue, courtroom data, pamphlets, and letters (etc.) to explore particular pragmatic/discoursal features within the EModE period. Some of these studies have adopted a “usage-based approach to language change”, which prioritizes the grammaticalization of linguistic elements over time (see, e.g., Traugott 2004: 538). Others have taken an approach that is more overtly sociolinguistic in orientation (see, e.g., the investigations relating to im/politeness). What I have not discussed in detail is the number of researchers who focus on multiple pragmatic/discoursal features within a given text-type/genre in order to say something meaningful about (the structure of) EModE discourse/genre(s) more generally – or, indeed, the work of a particular playwright. Length constraints
41 Early Modern English: Pragmatics and discourse prevent a detailed overview of all of the genres that have been studied to date. Readers are therefore encouraged to utilize the studies in Jucker et al. (eds.) (1999), in conjunction with Jucker’s bibliographical list of historical pragmatic studies (http://es-jucker. uzh.ch/HistPrag.htm, last accessed 14 September 2011), as a starting point for further investigation. Suffice it to say, there is a wealth of data exploring Shakespeare’s works: Rudanko (1993), for example, has utilized Grice’s (1975) cooperative principle, Brown and Levinson’s (1987 [1978]) politeness theory, and speech act theory in his studies of Othello, Coriolanus, and Timon of Athens. Magnusson (1999) has also drawn on Brown and Levinson (1987 [1978]), in conjunction with Bourdieu’s (1991) idea of the linguistic market, to explore the rhetoric of verbal interaction in Shakespeare. In addition, she compares Shakespeare’s works with those of his contemporaries (e.g., Erasmus). Taavitsainen is among a number of researchers who have investigated the dialogic elements of EModE medical/scientific writing; drawing on a corpus of medical treatises originally published between 1375–1750 (see, e.g., Taavitsainen and Pahta 1997). Taavitsainen (1999) shows how instruction in scientific handbooks was often given in the form of fictional conversations between a master and a pupil. Her approach involves combining close reading with computerized searches for elements with dialogic potential (see Jucker et al. 1999: 20). Correspondence is probably the most-researched genre in the EModE period, however. The bulk of the studies have benefited from the development of the Corpus of Early English Correspondence. Indeed, the latter has been used to study address terms, directives and salutations, personal pronouns, discourse markers, reporting, subjectivity and stance (see, e.g., Nevalainen and Raumolin-Brunberg 1995; Raumolin-Brunberg 1996; Fitzmaurice 2000). Many of the historical studies highlighted above are socio-pragmatic in orientation. For example, researchers who study correspondence regularly take account of issues relating not only to the production and distribution of letters, but also to the levels of literacy, relevant social distinctions (in particular, social status and gender), the conventional form of the letter and the possible impact of letter-writing manuals on the latter (see, e.g., Palander-Collin 2010). Some researchers have also begun to develop socio-pragmatic annotation schemes, which provide historically appropriate information respecting the status, role, age and gender of the interactants at the utterance level (see, e.g., Archer and Culpeper 2003), and have used these schemes to explain the discursive strategies of, for example, courtroom participants in the later EModE period (see, e.g., Archer 2005). This would suggest that an emerging trend, when looking at historical language use, is one which pays attention to the “interplay between the individual, language and society” (Palander-Collin 2010; see, also, Culpeper 2010); that is to say, to what is happening, linguistically, within a given text or speech event (i.e. the micro-interactional aspects), to what was happening around it (which necessitates our making use of both the insights of historians and also contemporary opinion/ideas), and, finally, to “how” and “why” (i.e., the purpose[s] for which) it was produced.
6 References Aijmer, Karin. 1996. Conversational Routines in English. London: Longman. Akimoto, Minoji. 2000. The grammaticalisation of the verb “pray”. In: Olga Fischer, Annette Rosenbach, and Dieter Stein (eds.), Pathways of Change: Grammaticalisation in English, 67–84. Amsterdam/Philadelphia: John Benjamins.
663
664
V Early Modern English Archer, Dawn. 2005. Questions and Answers in the English Courtroom (1640–1760). Amsterdam/ Philadelphia: John Benjamins. Archer, Dawn. 2008. Verbal aggression and impoliteness: Related or synonymous? In: Derek Bousfield and Miriam Locher (eds.), Impoliteness in Language: Studies on its Interplay with Power in Theory and Practice, 181–207. Berlin/New York: Mouton de Gruyter. Archer, Dawn and Jonathan Culpeper. 2003. Sociopragmatic annotation: New directions and possibilities in historical corpus linguistics. In: Andrew Wilson, Paul Rayson, and Tony McEnery (eds.), Corpus Linguistics by the Lune: A Festschrift for Geoffrey Leech, 37–58. Frankfurt: Peter Lang. Arnovick, Leslie K. 1994. The expanding discourse of promises in Present-Day English: A case study in historical pragmatics. Folia Linguistica Historica 15(1–2): 175–191. Arnovick, Leslie K. 1999. Diachronic Pragmatics. Seven Case Studies in English Illocutionary Development. Amsterdam/Philadelphia: John Benjamins. Atkinson, Dwight. 1999. Scientific Discourse in Sociohistorical Context: The Philosophical Transactions of the Royal Society of London, 1675–1975. Mahwah, NJ: Laurence Erlbaum. Austin, John L. 1962. How to Do Things with Words. Oxford: Oxford University Press. Blake, Norman F. 1992. Why and what in Shakespeare. In: Toshiyuki Takamiya and Richard Beale (eds.), Chaucer to Shakespeare: Essays in Honour of Shinsuke Ando, 179–193. Cambridge: D. S. Brewer. Blake, Norman F. 1996. Essays on Shakespeare’s Language. Misterton: The Language Press. Blum-Kulka, Shoshana and Juliane House. 1989. Cross-cultural and situational variation in requesting behavior. In: Blum-Kulka et al. (eds.), 123–154. Blum-Kulka, Shoshana, Juliane House, and Gabriele Kasper (eds.). 1989. Cross-cultural Pragmatics: Requests and Apologies. Norwood, NJ: Ablex Bonnelli, Elena T. 1992. “All I’m saying is …” : The correlation of form and function in pseudocleft sentences. Literary and Linguistic Computing 7(1): 30–42. Borgmeier, Raymond, Herbert Grabes, and Andreas H. Jucker (eds.). 1998. Anglistentag 1997 Giessen Proceedings. Trier: Wissenschaftlicher Verlag. Bourdieu, Pierre. 1991. Language and Symbolic Power. Cambridge: Polity Press. Bousfield, Derek. 2007. “Never a truer word said in jest”: A pragmastylistic analysis of impoliteness as banter in Henry IV, Part I”. In: Marina Lambrou and Peter Stockwell (eds.), Contemporary Stylistics, 210–220. London: Continuum. Bousfield, Derek. 2008. Impoliteness in Interaction. Amsterdam/Philadelphia: John Benjamins. Boyer, Abel. 1702. The English Theophrastus. London: W. Turner. Brinton, Laurel J. 1996. Pragmatic Markers in English. Berlin/New York: Mouton de Gruyter. Brinton, Laurel J. 2001. Historical discourse analysis. In: Deborah Schiffrin, Deborah Tannen, and Heidi E. Hamilton (eds.), The Handbook of Discourse Analysis, 138–160. Oxford: Blackwell. Brinton, Laurel. 2006. Pathways in the development of pragmatic markers in English. In: Ans van Kemenade and Bettelou Los (eds.), The Handbook of the History of English, 307–334. London: Blackwell. Brinton, Laurel J. 2010. Discourse markers. In: Jucker and Taavitsainen (eds.), 285–314. Brown, Penelope and Stephen C. Levinson. 1987 [1978]. Politeness. Some Universals in Language Usage. Cambridge: Cambridge University Press. Brown, Roger and Albert Gilman. 1960. The pronouns of power and solidarity. In: Thomas A. Sebeok (ed.), Style in Language, 253–276. Cambridge, MA: MIT Press. Brown, Roger and Albert Gilman. 1989. Politeness theory and Shakespeare’s four major tragedies. Language in Society 18(2): 159–212. Busse, Ulrich. 1999. “Prithee now, say you will, and go about it”. Prithee vs. pray you as discourse markers in the Shakespeare corpus. In: Fritz-Wilhelm Neumann and Sabine Schu¨lting (eds.), Anglistentag 1998 Erfurt Proceedings, 485–500. Trier: Wissenschaftlicher Verlag. Busse, Ulrich. 2002. Changing politeness strategies in English requests: A diachronic investigation. In: Jacek Fisiak (ed.), Studies in English Historical Linguistics and Philology: A Festschrift for Akio Oizumi, 17–35. Frankfurt/Main: Peter Lang.
41 Early Modern English: Pragmatics and discourse Busse, Ulrich. 2008. An inventory of directives in Shakespeare’s King Lear. In: Jucker and Taavitsainen (eds.), 85–114. Calvo, Carla. 1992. Pronouns of address and social negotiation in As You Like It. Language and Literature 1(1): 5–27. Claridge, Claudia. 2005. Questions in Early Modern English pamphlets. Journal of Historical Pragmatics 6(1): 133–168. Cole, Peter and Jerry Morgan (eds.). 1975. Syntax and Semantics 3: Speech Acts. New York: Academic Press. Comrie, Bernard. 1976. Linguistic Politeness Axes: Speaker-addressee, Speaker-referent, Speakerbystander. Pragmatics Microfiche 1.7: A3. Cambridge: Department of Linguistics, University of Cambridge. Culpeper, Jonathan. 1996. Towards an anatomy of impoliteness. Journal of Pragmatics 25: 349–367. Culpeper, Jonathan. 1998. (Im)politeness in drama. In: Peter Verdonk, Mick Short, and Jonathan Culpeper (eds.), Exploring the Language of Drama: From Text to Context, 83–95. London: Routledge. Culpeper, Jonathan. 2005. Impoliteness and The Weakest Link. Journal of Politeness Research 1(1): 35–72. Culpeper, Jonathan. 2010. Historical sociopragmatics. In: Jucker and Taavitsainen (eds.), 69–94. Culpeper, Jonathan and Dawn Archer. 2008. Requests and directness in Early Modern English trial proceedings and play texts, 1640–1760. In: Jucker and Taavitsainen, (eds.), 45–84. Culpeper, Jonathan and Merja Kyto¨. 1999. Modifying pragmatic force: Hedges in a corpus of Early Modern English dialogues. In: Jucker et al. (eds.), 293–312. Culpeper, Jonathan and Elena Semino. 2000. Constructing witches and spells: Speech acts and activity types in Early Modern England. Journal of Historical Pragmatics 1(1): 97–116. Culpeper, Jonathan, Derek Bousfield, and Anne Wichmann. 2003. Impoliteness revisited: With special reference to dynamic and prosodic aspects. Journal of Pragmatics 35(10–11): 1545–1579. Edmonson, Willis J. and Juliane House. 1981. Let’s Talk and Talk about it: A Pedagogic Interactional Grammar of English. Munich: Urban & Schwarzenberg. Fairclough, Norman F. 1993. Discourse and Social Change. Cambridge: Polity Press. Fischer, Andreas. 1998. Marry: From religious invocation to discourse marker. In: Borgmeier et al. (eds.), 35–46. Fitzmaurice, Susan. 2000. Tentativeness and insistence in the expression of politeness in Margaret Cavendish’s Sociable Letters. Language and Literature 9(1): 7–24. Fraser, Bruce. 1999. What are discourse markers? Journal of Pragmatics 31: 931–952. Goffman, Erving. 1967. Interaction Ritual. Chicago: Aldine. Go¨rlach, Manfred. 1991. Introduction to Early Modern English. Cambridge: Cambridge University Press. Grice, H. Paul. 1975. Logic and conversation. In: Cole and Morgan (eds.), 41–58. Hope, Jonathan. 1994. The use of thou and you in Early Modern spoken English: Evidence from depositions in the Durham Ecclesiastical Court Records. In: Dieter Kastovsky (ed.), Studies in Early Modern English, 141–152. Berlin/New York: Mouton de Gruyter. Jacobs, Andreas and Andreas H Jucker. 1995. The historical perspective in pragmatics. In: Jucker (ed.), 3–33. Jacobsson, Mattias. 2002. Thank you and thanks in Early Modern English. ICAME Journal 26: 63–80. Jucker, Andreas H. 1997. The discourse marker well in the history of English. English Language and Linguistics 1(1): 91–110. Jucker, Andreas H., Gerd Fritz, and Franz Lebsanft. 1999. Historical dialogue analysis: Roots and traditions in the study of the Romance languages, German and English. In: Jucker et al. (eds.), 1–34. Jucker, Andreas H. 2000. History of English and English Historical Linguistics. Stuttgart: Klett. Jucker, Andreas H. 2002. Discourse markers in Early Modern English. In: Richard J. Watts and Peter Trudgill (eds.), Alternative Histories of English, 210–230. London: Routledge.
665
666
V Early Modern English Jucker, Andreas H. (ed.). 1995. Historical Pragmatics. Pragmatic Developments in the History of English. Amsterdam/Philadelphia: John Benjamins. Jucker, Andreas H. and Irma Taavitsainen. 2000. Diachronic speech act analysis: Insults from flyting to flaming. Journal of Historical Pragmatics 1(1): 67–95. Jucker, Andreas H. and Irma Taavitsainen. 2003. Diachronic perspectives on address term systems: introduction. In: Irma Taavitsainen and Andreas H. Jucker. (eds.), Diachronic Perspectives on Address Term Systems, 1–26. Amsterdam/Philadelphia: John Benjamins. Jucker, Andreas H. and Irma Taavitsainen. 2008. Apologies in the history of English: Routinized and lexicalized expressions of responsibility and regret. In: Jucker and Taavitsainen (eds.), 229–244. Jucker, Andreas H., Gerd Fritz, and Franz Lebsanft (eds.). 1999. Historical Dialogue Analysis. Amsterdam/Philadelphia: John Benjamins. Jucker, Andreas H. and Irma Taavitsainen (eds.). 2008. Speech Acts in the History of English. Amsterdam/Philadelphia: John Benjamins. Jucker, Andreas H. and Irma Taavitsainen (eds.). 2010. Handbook of Historical Pragmatics. Berlin/ New York: De Gruyter Moutor. Klein, Lawrence E. 1994a. “Politeness” as linguistic ideology in late seventeenth- and eighteenthcentury England. In: Dieter Stein and Ingrid Tieken-Boon van Ostade (eds.), Towards a Standard English 1600–1800, 31–50. Berlin/New York: Mouton de Gruyter. Klein, Lawrence E. 1994b. Shaftesbury and the Culture of Politeness: Moral Discourse and Cultural Politics in Early Eighteenth-century England. Cambridge: Cambridge University Press. Kohnen, Thomas. 2002. Methodological problems in corpus based historical pragmatics: The case of English directives. In: Karin Ajimer and Bengt Altenberg (eds.), Language and Computers. Advances in Corpus Linguistics. Papers from the 23rd International Conference on English Language Research on Computerized Corpora (ICAME 23). Go¨teborg, 22–26 May, 237–247. Amsterdam: Rodopi. Kohnen, Thomas. 2004. “Let mee bee so bold to request you to tell mee”: Constructions with “let me” and the history of English directives. Journal of Historical Pragmatics 5(1): 159–173. Kopytko, Roman. 1993. Polite Discourse in Shakespeare’s English. Poznan´: Wydawnictwo Naukowe Uniwersytetu im. Adam Mickiewcza w Poznaniu. Kopytko, Roman. 1995. Linguistic politeness strategies in Shakespeare’s plays. In: Jucker (ed.), 515–540. Kryk-Kastovsky, Barbara. 1998. Pragmatic particles in Early Modern English court trials. In: Borgmeier et al. (eds.), 45–56. Kryk-Kastovsky, Barbara. 2006. Impoliteness in Early Modern English courtroom discourse. Journal of Historical Pragmatics 7(2): 213–244. Leech, Geoffrey N. 1983. Principles of Pragmatics. London: Longman. Levinson, Stephen C. 1983. Pragmatics. Cambridge: Cambridge University Press. Magnusson, Lynne. 1999. Shakespeare and Social Dialogue: Dramatic Language and Elizabethan Letters. Cambridge: Cambridge University Press. Mills, Sarah. 2003. Gender and Politeness. Cambridge: Cambridge University Press. Nevala, Minna. 2004. Accessing politeness axes: Forms of address and terms of reference in early English correspondence. Journal of Pragmatics 36: 2125–2160. Nevalainen, Terttu and Helena Raumolin-Brunberg. 1995. Constraints on politeness: The pragmatics of address formulae in Early English correspondence. In: Jucker (ed.), 541–601. Nikula, Tarja. 1996. Pragmatic Force Markers: A Study in Interlanguage Pragmatics. Studia Philologica Jyva¨skyla¨ensia: University of Jyva¨skyla¨. Palander-Collin, M. 2010. Correspondence. In: Jucker and Taavitsainen (eds.), 651–677. Raumolin-Brunberg, Helena. 1996. Forms of address in Early English correspondence. In: Terttu Nevalainen and Helena Raumolin-Brunberg (eds.), Sociolinguistics and Language History: Studies based on the Corpus of Early English Correspondence, 167–181. Amsterdam: Rodopi.
41 Early Modern English: Pragmatics and discourse
667
Rudanko, Juhani. 1993. Pragmatic Approaches to Shakespeare. Essays on Othello, Coriolanus and Timon of Athens. Lanham: University Press of America. Rudanko, Juhani. 2006. Aggravated impoliteness and two types of speaker intention in an episode in Shakespeare’s Timon of Athens. Journal of Pragmatics 38(6): 829–841. Schlieben-Lange, Brigitte. 1976. Fu¨r einer historische Analyse von Sprechakten. In: Heinrich Weber and Harald Weydt (eds.), Sprachtheorie und Pragmatik, Akten des 10. Linguistischen Kolloquiums Tu¨bingen 1975, 113–119. Tu¨bingen: Niemeyer. Schlieben-Lange, Brigitte. 1983. Traditionen des Sprechens. Stuttgart: Kohlhammer. Schlieben-Lange, Brigitte and Heinrich Weydt. 1979. Streitgespra¨ch zur historita¨t von Sprechakten. Linguistiche Berichte 60: 65–78. Schwenter, Scott A. and Elizabeth C. Traugott. 2000. Invoking scalarity: The development of in fact. Journal of Historical Pragmatics 1: 7–25. Searle, John R. 1969. Speech Acts: An Essay in the Philosophy of Language. Cambridge: Cambridge University Press. Searle, John R. 1975. Indirect speech acts. In: Cole and Morgan (eds.), 59–82. Taavitsainen, Irma. 1995. Interjections in Early Modern English: From initiation of spoken to conventions of written language. In: Jucker (ed.), 439–465. Taavitsainen, Irma. 1999. Dialogues in Late Medieval and Early Modern English medical writing. In: Jucker, Fritz, and Lebsanft (eds.), 243–268. Taavitsainen, Irma and Andreas H. Jucker. 2008a. Speech acts now and then: Towards a pragmatic history of English. In: Jucker and Taavitsainen (eds.), 1–23. Taavitsainen, Irma and Andreas H. Jucker. 2008b. “Methinks you seem more beautiful than ever”: Compliments and gender in the history of English. In: Jucker and Taavitsainen (eds.), 195–228. Taavitsainen, Irma and Pa¨ivi Pahta. 1997. The Corpus of Early English Medical Writing. ICAME Journal 21: 71–78. Traugott, Elizabeth Closs. 1982. From propositional to textual and expressive meanings: Some semantic-pragmatic aspects of grammaticalization. In: Winifred P. Lehmann and Yakov Malkiel (eds.), Perspectives on Historical Linguistics, 245–271. Amsterdam/Philadelphia: John Benjamins. Traugott, Elizabeth Closs. 1995. The role of the development of discourse markers in a theory of grammaticalization. Paper presented at the Twelfth International Conference on Historical Linguistics, Manchester, August 1995. http://www.stanford.edu/~traugott/ect-papersonline.html Traugott, Elizabeth Closs. 2004. Historical pragmatics. In: Laurence Horn and Gregory Ward (eds.), The Handbook of Pragmatics. Oxford: Blackwell, 538–561. Traugott, Elizabeth C. 2008. “All that he endeavoured to prove was …”: On the emergence of grammatical constructions in dialogual and dialogic contexts. In: Robin Cooper and Ruth Kempson (eds.), Language in Flux: Dialogue Coordination, Language Variation, Change and Evolution, 143–177. London: Kings College Publication. Traugott, Elizabeth C. and Richard Dasher. 2002. Regularity in Semantic Change. Cambridge: Cambridge University Press. Tsui, Amy B. M. 1994. English Conversation. Oxford: Oxford University Press. Valkonen, Petteri. 2008. Showing a little promise: Identifying and retrieving explicit illocutionary acts from a corpus of written prose. In: Jucker and Taavitsainen (eds.), 247–272. Walker, Terry. 2007. Thou and you in Early Modern English Dialogues: Trials, Depositions and Drama Comedy. Amsterdam/Philadelphia: John Benjamins. Watts, Richard J. 2003. Politeness. Cambridge: Cambridge University Press. Wierzbicka, Anna. 1987. English Speech Act Verbs: A Semantic Dictionary. Sydney: Academic Press.
Dawn Archer, Preston (UK)
668
V Early Modern English
42 Early Modern English: Dialects 1 2 3 4 5
The concept of dialect Sources for the study of regional variation in Early Modern English Linguistic systems and subsystems illustrating dialectal variation Further research References
Abstract In comparison with both Middle English and Late Modern, and Present-day English, regional variation in the Early Modern English period has been a less intensively researched area owing to the lack of quantitatively and qualitatively valid data in digital form. Instead of the data-driven and data-oriented approach, historical dialectology focusing on this period has often yielded to language attitudes and the prescriptivist tradition, defining dialect use as non-standard and describing it with reference to marked features exclusively. In the 21st century, several new corpora have been compiled to increase the diatopic representativeness of EModE databases, those drawing on manuscripts being particularly relevant. In modern corpus-based dialect research, the description of linguistic systems at the local and regional levels is based on comprehensive inventories of all features in a particular area, not a pre-selected set of features, probably biased or prejudiced, alleged to permit a diagnosis of a particular variety as dialectal.
1 The concept of dialect In English historical linguistics, dialects have been examined from a wide range of perspectives. These seem to reflect the prevailing theoretical and methodological approaches adopted in research on particular periods in the history of English, for example, the focus on standardization distinguishing dialectology in the Early Modern English period from that which draws on data from the earlier time-periods. The main claim in this chapter is that dialectology is rarely about diatopic variation exclusively. Research on dialects, in Modern and Present-day English in particular, tends instead to discuss attested variation in a particular region or locality referring to a wide range of variables, chiefly sociolinguistically defined. The diverse approaches to the discipline can be illustrated for example by the following four themes dialectologists have been concerned about. Firstly, dialects have been considered as reified entities (Milroy 1999; Benson 2001). As pointed out in Meurman-Solin (2004): There is a tendency to objectify or reify regional varieties by assuming they form relatively homogeneous – perhaps even relatively self-contained – entities or systems, a tendency to historicise by stressing socio-political rather than linguistic factors as legitimising the naming and describing of regional varieties in a particular way, and a tendency to hierarchicise, leading to the analysis of a regional variety with exclusive reference to a standardised variety (Meurman-Solin 2004: 27).
Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 668–685
42 Early Modern English: Dialects In this approach, the debate is often about whether a particular variety is a dialect or in fact a language, the case in point being the debated status of Scottish English in the Early Modern English period. This is a period in which a systematic preference for chiefly or uniquely Scottish features has been recorded, patterns of variation and change also, reflecting a sufficiently high degree of variety-internal uniformity to permit the use of the term “Scottish Standard” (Devitt 1989; Meurman-Solin 1993; see also Go¨rlach 1996: 468–470; Dossena 2005: 37–55). Secondly, regional variation has been studied from an explicitly sociolinguistic perspective, highlighting the pace and direction of change, the diffusion of innovative variants, in particular as conditioned by variables such as the informant’s age, social rank, geographical and social mobility, and education. Among recent research, considerable advances in historical sociolinguistics have been made possible by the team involved in creating the family of the Early English Correspondence corpora (CEEC) (Nevalainen et al. 1998) (e.g. Nevalainen and Raumolin-Brunberg 1996, 2003; see Raumolin-Brunberg, Chapter 45; Nevalainen, Volume 2, Chapter 92). In these studies, “region” is just one of the many variables affecting how the CEEC family of corpora have been compiled and how long-term diachronic developments recorded in the data have been interpreted. Thirdly, the approach adopted in the creation of linguistic atlases at the Institute for Historical Dialectology (IHD), University of Edinburgh, is to investigate diachronic, diatopic and diastratic variation. However, because the main goal is to map variation and change (i.e. to present the areal diffusion as maps of various kinds as illustrated in Williamson 2004: 126–128), theoretical thinking and the creation of sophisticated tools for data retrieval and analysis have focused on how to conceptualize the variables of time and space in synchrony and diachrony and to localize texts by the so-called “fitting technique” (on this pioneering work in historical dialectology at the IHD, see, for example, Laing 2004; Williamson 2004; and Laing and Lass 2006; for information on current projects of the IHD, see http://www.lel.ed.ac.uk/research/ihd/projectsX.shtml). It is noteworthy, however, that methods as well as principles and practices of corpus compilation tailored for periods such as Middle English and Older Scots are not directly transferable to the study of Early Modern English dialects but will have to be developed further for periods in which distinguishing the conditioning of time and space in isolation from that of other variables may not be possible. With the introduction of printing, the preferred practices of printing houses as well as the perceived wider appeal of standardized varieties meant that writing in a particular dialect was restricted to a limited range of genres. Consequently, explicitly defined criteria will have to be formulated to identify texts that can be claimed to represent a particular dialectal variety. Fourthly, in the perhaps most widespread approach to dialects and dialectology, research in the field has invited the introduction of language attitudes. This suggests that dialectal features are – implicitly or explicitly – conceptualized as marked, or even stigmatized. The term type which reflects the comparative angle is also widespread: in contrast with standard varieties, dialects are non-standard or sub-standard, even though features cited as representing a particular dialect may be basic to a particular local or regional standard. Thus Go¨rlach (1999) distinguishes between regional and local varieties by comparison with supraregional standards, focusing in his chapter on illustrating contemporary attitudes to the former, and commenting on how prescriptivist trends are reflected in grammars and word-lists recommending good usage and
669
670
V Early Modern English the avoidance of “non-standard” features (cf. Dossena 2005 on Scotticisms). Instead of drawing on texts that can be claimed to represent a particular dialectal variety by language-external criteria, Go¨rlach quotes comments on distinctively dialectal language use in contemporary, or sometimes later literature, or varieties constructed for literary purposes, such as Stage Irish. Research on dialects tends to focus on orthography, phonology and lexis. Go¨rlach points out that Of all the linguistic levels, it is probably most difficult to distinguish between regional and social non-standard features in syntax, since both would have been levelled away by the prescriptive influence of the schools in the surviving written texts. Moreover, syntactic variation is more often a consequence of stylistic choice, depending on formality, text type and topic: where one writer may deliberately flout the rules of ‘correct’ syntax by using conversational style (in a private letter or in a personal diary not intended for anyone else to read), another may use simple syntax because he cannot do any better (Go¨rlach 1999: 491).
The present chapter claims that the concept of dialectal variation should not be viewed as related to “non-standard features”, nor is it appropriate to relate dialectal language use to stylistic choice. The passage quoted above does however illustrate the confusing ways of using the concept of dialect, the reference to register variation (“conversational style”) being quite inappropriate in this context (cf. the discussion of stylistic competence in Johnston 1997: 51). In contrast with what seems to be implied by these remarks on style, private documents such as letters scarcely reflect conversational style, since, in early correspondence, writers tended to borrow polite formulae and use the quite formal negative politeness strategies considered appropriate in this particular communicative function in contemporary society. As regards spelling variation, data extracted from corpora do not justify Go¨rlach’s view that “the occasional mis-spellings that are found in later Early Modern English private documents are generally diagnostic of social status rather than of dialect” (Go¨rlach 1999: 506). In fact, spelling variation in letters has been shown to reflect local and regional practices (Meurman-Solin 2000a, 2001). In the present approach, the term “dialect” is used to refer to language use attested in a particular geographical area, the data comprising both supraregional and regional or local features. In other words, instead of concentrating on distinctive or marked features exclusively, language use in a particular region is described by using a comprehensive inventory. The general assumption is that at least some degree of systematicity as regards distinctive preferences at the local or regional level can be identified as well as some degree of influence by contacts with other local or regional varieties and supraregional standardized varieties. Thus my concern here is not to identify what is chiefly or exclusively local or regional but to provide a full account of linguistic systems drawing on a representative corpus of texts originating from a particular locality or region in a particular time-period. Recently developed methods of historical dialectology will be illustrated here by drawing on digital corpora representing some southern and northern English varieties in the 16th and 17th centuries. The approach to Scottish English here is to view language use in particular regions in Scotland as reflecting variation and change in English, leaving aside the debate about whether Scots is an independent language or not (cf. Go¨rlach 1996, 1999: 468–469). It will also become clear that I prefer to
42 Early Modern English: Dialects make a distinction between the theoretical and methodological approach of historical sociolinguistics (as in Nevalainen and Raumolin-Brunberg 2003) and that of historical dialectology. These approaches are often merged in various ways (e.g. Go¨rlach 1999: 470–474), making it difficult to understand mechanisms of language variation and change conditioned by geographical space. Since I find it necessary to define the domain of historical dialectology in stricter terms, its main task is to “describe” in detail language use over time and space. Only in the attempt to “interpret” variation and change attested in the data is it legitimate to start enquiring whether there are sociolinguistically relevant variables which correlate with the frequencies and distributions of linguistic features.
2 Sources for the study of regional variation in Early Modern English 2.1 New corpora Starting from the 15th century, written English represents local or regional language use quite unevenly due to the diffusion of a new standard norm constituting a supraregional variety of English. The range of writing using local or regional varieties became more restricted in the sense that texts representing some genres were viewed as appealing to wider audiences, their printing in particular strengthening the need to use the supraregional standard. Scottish English became anglicized during the 17th century (see Devitt 1989; Meurman-Solin 1993, 1997); according to Go¨rlach (1999: 468), “Wales was still predominantly Welsh-speaking in the nineteenth century, and Cornish survived until the eighteenth.” At present, there is an increased awareness of a scarcity of research on regional variation in the Early Modern English period due to lack of quantitatively and qualitatively valid data. The most neglected area is probably the investigation of manuscript sources in the various record offices and archives. These relevant data sources should be transcribed and digitized to make them internationally available. In selecting texts for a corpus, language-external criteria should be applied. Since most existing corpora supply a wide variety of both literary and non-literary texts representing the range of genres traditionally included in general-purpose corpora, the focus in complementing these corpora should be on genres which have a local or regional communicative function. The new corpora or new supplements to existing ones would include legal texts of various kinds produced by the local administrative and legal institutions as well as private transactions such as land transfers arranged by advocates, official correspondence, and a wide range of private writing, diaries, autobiographies, accounts, commonplace books, and letters in particular. The linguistic examination of these will inevitably show that there are quite complex texts among these; some texts, such as trial proceedings consist of a number of different texts, witness depositions often reflecting local linguistic practices, while some other texts represent legalese, for example, formulaic language use adapting Latin or French models. It is generally agreed that historical dialectology in the Early Modern English period is not as advanced as that focusing on Middle English. While the Linguistic Atlas of Late Middle English (LALME) (McIntosh et al. 1986) and especially the recently launched Linguistic Atlas of Early Middle English (LAEME) (Laing and Lass 2007)
671
672
V Early Modern English have completely changed our ability to reconstruct dialectal variation in Middle English, the situation is quite different in Early Modern English, as Kyto¨, Grund, and Walker (2007) point out: We still know comparatively little about regional Early Modern Englishes: they have been difficult to study since edited material that can be clearly anchored in a specific region of England is scarce. The ongoing standardisation of English during the period also means that regional features were being increasingly suppressed in favour of the norm in most formal written records (Kyto¨, Grund, and Walker 2007: Section 4).
As regards the availability of relevant data, innovative work by Kyto¨ et al. (Kyto¨, Walker, and Grund 2007; see also Kyto¨, Grund, and Walker 2007) has improved the situation by producing a digital edition of manuscripts of English witness depositions dating from 1560–1760, representing four regions of England (the North, South, East, and West) and the London area. As they rightly claim, “[w]ritten records of the oral testimony of a witness in a criminal or ecclesiastical trial give a glimpse into the lives and language of ordinary people in the past”. However, they also stress that, even though in principle witness depositions could be considered an ideal source for investigating regional variation, there are caveats. Kyto¨, Grund, and Walker (2007) list the following: since legal writing is a very formal genre, standardisation is only to be expected, and so is perhaps even the removal of “inappropriate” regional language. Although the presence of dialectal vocabulary does show that all regionalisms were not sifted out by the scribe, it is frequently difficult to determine whether the language reported by the scribe is indeed that of the witness: the language may in fact be partly or wholly that of the scribe (see Kyto¨ and Walker 2003; Grund, Kyto¨, and Rissanen 2004; Grund 2007). The level of training of the scribe probably contributed to the picture: the more highly educated would perhaps have been more discriminating, while less trained scribes may have followed the actual language of the deponents more closely. Unfortunately, there is little information on the scribes, and more studies on the Early Modern legal scribes that worked for justices of the peace and different courts remain a desideratum (Kyto¨, Grund, and Walker 2007: Section 4).
However, Huber (2007) considerably increases our understanding of the practices of scribes by discussing those involved in the production of the Old Bailey Proceedings. At present, regional variation has been described in studies drawing on different data sources, but, most usefully, this fragmentary evidence is complementary, as Kyto¨, Grund, and Walker (2007) illustrates. While the CEEC (Nevalainen et al. 1998) corpora contain data on the North, East Anglia and the London area, the witness depositions originate from the counties of Northumberland, Cumberland, Westmorland, Yorkshire, Lancashire, Norfolk, Suffolk, Essex, Oxfordshire and Somerset in addition to the London area. The compilation and annotation principles of the manuscript-based Corpus of Scottish Correspondence (CSC) (Meurman-Solin 2007a), 1500–1715, have been thoroughly affected by those applied to the Linguistic Atlas of Early Middle English and the Linguistic Atlas of Older Scots (see the online introductions to LAEME [Laing and Lass 2007] and LAOS [Williamson 2007], and Meurman-Solin 2007a, 2007b). One of the goals of the CSC corpus is to be diatopically representative. In the 2007 version that the illustrations in Section 3 are based on, the dialect areas
42 Early Modern English: Dialects represented are as follows: Ross and Cromarty, Sutherland, Aberdeenshire, Angus, Perthshire, Fife, Lothian, the Border counties, Lanarkshire, Ayrshire, Argyllshire and Dumfries and Galloway. Following the principles applied to the Scottish National Dictionary (Grant and Murison 1931–76) and the Concise Scots Dictionary (Robinson 1985–96), these areas have been grouped into six regions: Northern Scots, North-East Scots, East Mid Scots, West Mid Scots, South Mid Scots, and Southern Scots. South Mid Scots and Southern Scots have mostly been ignored in presenting some illustrative data in Section 3, since samples representing them are not yet of a statistically valid size.
2.2 Assessment of data sources for historical dialectology The validity of findings in research on regional variation in Early Modern English is drastically reduced by the frequently stated fact that the data only represents the literate, i.e. mostly members of the higher social ranks (see Nevalainen and Raumolin-Brunberg 2003; on Scottish women’s literacy, see Meurman-Solin 2001: 21–22). Two factors in particular diminish our ability to provide a balanced reconstruction of historical varieties of English used in different localities. Firstly, there is as yet no purpose-built database enabling a thorough investigation of how rural and urban varieties relate to one another in the various regions and localities in the Early Modern English period. Beside the cline from local to regional and supraregional, there are other clines; for example, from peripherally located rural areas to rural areas within easy access in the network of transport and communication, and from economically self-supporting and culturally self-contained towns to lively royal boroughs favoring trade, with further enriching dimensions such as a university and important administrative and legal institutions. Secondly, social, cultural, and economic distance may play a more important role in linguistic variation and change than mere geographical distance. My earlier corpusbased research suggests that “distance” as a social, economic, and cultural construct, rather than as a concept defined purely geographically, is a significant conditioning factor in patterns of variation and change attested in the history of Scots (Meurman-Solin 2000a, b). In fact, in studying local and regional variation, it is necessary to define quite complex variables as factors conditioning linguistic choice. While the scalar concept of distance permits us to define relations in terms of core and periphery, for example, with respect to economic significance, the concept of network allows the assessment of type, direction and frequency of ties. In principle, all texts originating from a particular area can be claimed to represent it. However, their representativeness varies, there being a set of variables by which the texts can be positioned on a cline. As discussed in Meurman-Solin (2004, 2007a), texts written by members of a speech community are diatopically more representative than those by members of discourse and text communities. To comment on the latter two, members of discourse communities share a wide range of conventionalized discourse properties with geographically distant communities. Since there are important differences between what texts, genres and text types were available and circulated in a particular locality or region in a particular time-period, information about text community type also permits us to create criteria for assessing the validity and relevance of data for the study of dialectal variation. Since all the sources represent written language, fully
673
674
V Early Modern English authentic data that could be claimed to reflect spoken practices is rare. Letters by privately trained and inexperienced writers are often mentioned as the best source of local language use but the potential influence of models must be taken into account even in these. Moreover, the relatively low level of linguistic and stylistic competence of these writers may increase the number of hapax variants, the attested variants being idiosyncratic rather than diagnostic of a particular geographical area (see Meurman-Solin 1999, 2001, 2004, 2005). As with Old English and Middle English material, information about text histories will permit the quality assessment of data sources used for historical dialectology in the Early Modern English period. Texts may exist in different versions, their relation requiring detailed study. Printing may have had a major effect, standardizing being part of most printers’ policies. Texts spoken in a particular variety may have been written down by a scribe using a different variety. In principle, only diplomatic digital editions of manuscripts can be considered reliable, there being ample evidence of editorial practices applied in later centuries (including the present one) having permitted major processes of modernisation or normalisation (for standards in philological editing, see Lass 2004). Depending on how much language-external information is available and how reliable it is, texts can be positioned on a cline from primary witnesses, or “anchor texts”, to texts the status of which is unclear (cf. Nevalainen and Raumolin-Brunberg 1996: 43). A database must also be assessed from the perspective of whether it is valid for answering a particular research question, lexical studies not necessarily requiring the same standards of philological editing as phonological and morphosyntactic questions. In general, the relevance and validity of a corpus depends on which particular feature is investigated. Language-external variables may condition the general frequency of a feature, so that a quantitative analysis aiming to identify diatopically relevant patterns of variation may provide misleading results. For example, 45% of the instances of the personal pronoun SHE have been recorded in women’s letters in the CSC (Meurman-Solin 2007a), even though the proportion of these letters in the corpus is only circa 20%. Assessing the relative weight of region or locality and other variables defined by the sociolinguistic framework is a necessary but often difficult exercise.
3 Linguistic systems and subsystems illustrating dialectal variation Limitations of space prevent me from giving a full account of how the methodological approaches in recent historical dialectology succeed in depicting a particular linguistic system. This section will thus only provide some illustrations of patterns of diffusion across space using material from the CSC (Meurman-Solin 2007a) and CEEC (Nevalainen et al. 1998) corpora. The goal is merely to illustrate corpus-based analyses of diachronic developments in a number of regional varieties, and the findings based exclusively on letters have of course no validity across genres. In the letter genre, however, there are numerous representatives of local language use that cannot be recovered from among representatives of other genres. The linguistic features in the illustrations have been selected to permit comparison with data extracted from the CEEC (Nevalainen et al. 1998), discussed extensively in the publications of the compilers.
42 Early Modern English: Dialects
675
3.1 The relative system The relative system is an obvious choice for illustrating the theoretical and methodological approach of present-day historical dialectology, seeing that there is a long history of research in this area in Scottish studies (e.g. Caldwell 1974; Romaine 1982; Devitt 1989; Meurman-Solin 2000b, 2007c; see also Busse, Chapter 46; Johansson, Chapter 49).
3.1.1 Variant forms of which Formal variation in the relative which in the CSC (Meurman-Solin 2007a) can be traced by examining the variants as classified into three main types: the quhilk type, the transitional types quhich and whilk, here grouped together, and the which type. In this analysis, no distinction is made between occurrences in restrictive and non-restrictive relative clauses, the restrictive relative clauses introduced by which being quite infrequent in the CSC corpus (9% of the total of occurrences of which). Since the approach here is purely comparative, Figure 42.1 gives the findings in percentages. In the statistics, the quhilk type includes variants of quhilkis and the quhilk, and the which and quhich/whilk types include variants of the which and the quhich/whilk respectively. In the 16th century, all occurrences of this relative represent the quhilk type (there is no 16th-century data representing Northern Scots in the CSC). We can therefore focus on developments in four regions in the 17th century. 100%
80%
60% which quhich/whilk quhilk
40%
1600 –1649
Northern
West Mid
North East
East Mid
Northern
West Mid
North East
0%
East Mid
20%
1650 –1715
Figure 42.1: Percentages of the quhilk, quhich /whilk and which types in four regions in the CSC
As illustrated by Figure 42.1, all occurrences in the period 1650–1715 represent the which type in East Mid and Northern Scots, whereas the proportion of the quhilk type is still 10% and 2%, and that of quhich/whilk 20% and 3% in North East and West Mid Scots respectively. The quite large proportion of transitional variants in North East Scots as late as post-1650 letters is a particularly interesting finding.
676
V Early Modern English As regards the period 1600–1649, in which variation and change is generally most intense in the history of Scottish English, the proportion of the quhilk type is lowest in West Mid Scots (18%) and highest in the North (83%) and the North East (55%). However, the percentage is also quite high in East Mid Scots (48%). The high degree of variation attested does not exclusively result from quhilk variants being replaced by which variants. This can be illustrated by the variational pattern recorded in West Mid Scots, the percentage of the quhich/whilk type, i.e. the transitional variants, being as high as 28%, with the quhilk type representing 18% and the which type 54%. The proportion of the transitional variants is lowest in the North (1%) but considerably higher in East Mid and North East Scots (9% and 20% respectively). Overall, the variational patterns of the forms of the relative which resemble one another in East Mid and North East Scots in comparison with the quite different profiles of West Mid and Northern Scots. The main difference between the two groups is the greater number of transitional variants in the West and a clear time-lag in the spread of the which type recorded in Northern Scots. Moreover, only very few transitional variants have been attested in Northern Scots. This investigation has identified clear differences in the pace at which incoming variants spread across regions rather than demonstrating the existence of distinctive variants which are diagnostic of particular regional varieties. In comparing processes of diffusion, regions also differ from one another in the number and range of transitional variants. Figure 42.2 illustrates the proportion of transitional variants in the four regions (the type has not been attested in the 16th century): 50 East Mid Scots North East Scots West Mid Scots Northern Scots
40
30
20
10
0 1600–1649
1650–1715
Figure 42.2: Percentages of the quhich/whilk type in four regions in the 17th century in the CSC
While the relative proportion of quhich/whilk decreases to 3% in West Mid Scots and maintains the same level in North East Scots (20%), the type disappears from post-1650 letters representing East Mid and Northern Scots. As regards the minority variants of quhilk and which, the plural variant quhilkis is considerably more frequent in East and West Mid Scots than North East Scots. Nevalainen and Raumolin-Brunberg (2000) have studied variation between which
42 Early Modern English: Dialects
677
and the which, drawing on data extracted from the Corpus of Early English Correspondence. Figure 42.3 is based on information in Table 10 in Nevalainen and Raumolin-Brunberg (2000: 320): 70 60 London East Anglia North
50 40 30 20 10 0 1460–1499
1500 –1539
1540 –1579
1580 –1619
1620–1659
1660 –1681
Figure 42.3: Percentages of the which type in the three dialect areas in the CEEC
We notice that there is a rather abrupt change in the proportion of the the which type in the CEEC (Nevalainen et al. 1998) data, developments in the English and Scottish regions resembling one another from 1500 onwards (Figure 42.4). 70 60 East Mid Scots North East Scots West Mid Scots Northern Scots
50 40 30 20 10 0 1500–1549
1550 –1599
1600 –1649
1650 –1715
Figure 42.4: Percentages of the quhilk/which in four regions in the CSC
In the Scots data, a clearly larger proportion of the quhilk/which has been recorded in West Mid Scots in the first half of the 16th century and a regained 10% frequency in East Mid Scots in the first half of the 17th; otherwise there is a constant decrease in the use of this variant.
678
V Early Modern English
3.1.2 Relative that, which, who and a zero link in four regions of Scotland Due to limitations of space, only some patterns in the relative system can be illustrated here. It is possible to examine their diatopic distribution in greater detail since zero realizations of relative links have been signalled in the annotation applied to the CSC (see Meurman-Solin 2007a). Figure 42.5 illustrates zero relatives in the function of object with an inanimate antecedent, the most frequent function in which zero realizations occur: 100 East Mid Scots North East Scots West Mid Scots Northern Scots
80
60
40
20
0 1500 –1549
1550 –1599
1600 –1649
1650 –1715
Figure 42.5: Percentages of zero relatives in the function of object with an inanimate antecedent in four regions of Scotland in the CSC
Figure 42.5 displays a highly interesting similarity between developments in all the regions except West Mid Scots, where the general direction of change finally becomes convergent with the other regions at the beginning of the 17th century. A zero realization is attested in the range of 60–86% in 16th-century West Mid Scots. Figure 42.6 illustrates variation in the choice of a relative link in East Mid Scots: The analysis identifies a considerable increase in the proportion of zero realizations and a decrease in the proportion of the relative that in East Mid Scots. Of course, a quantitative result of this kind only provides one perspective on the description of the relative system in East Mid Scots.
3.2 Subject–verb agreement The second case study considers subject–verb concord in the present indicative and also presents some quantitative analyses on variation between s-variants, thvariants and zero-variants in the CSC (Meurman-Solin 2007a). This feature was selected since Nevalainen and Raumolin-Brunberg, the leaders of the CEEC project, have discussed it in great detail in their publications (see also Meurman-Solin 1992).
42 Early Modern English: Dialects
679
100 zero that which
80
60
40
20
0 1500–1549
1550 –1599
1600 –1649
1650 –1715
Figure 42.6: Percentages of the realizations of a relative link in the function of object with an inanimate antecedent in East Mid Scots in the CSC
3.2.1 -s versus zero in the first person singular present indicative The main perspective is that of tracing the degree of homogeneity in the application of the Northern Subject–Verb rule. Figure 42.7 illustrates the diffusion of s-variants in the first person present indicative with a non-adjacent pronoun subject: 100
80
60
Northern North East East Mid West Mid Southern
40
20
0 1500–1549
1550 –1599
1600 –1649
1650 –1715
Figure 42.7: Percentages of s-variants in the first person singular present indicative with a nonadjacent pronoun subject in the CSC
Figure 42.7 shows that the Northern Subject–Verb rule is most consistently used in East Mid Scots, the heartland of the so-called “Scottish Standard” in the 16th century and the first half of the 17th, the proportion of the s-variant remaining as large as 71% as late as the last sub-period. In at least 50% of the occurrences, the s-variant is also preferred throughout in West Mid and North East Scots, this variant being chosen in
680
V Early Modern English 80–100% of the cases in these areas in the 16th century. It is noteworthy that the rule loses its strength first in Northern Scots, the most peripheral area.
3.2.2 -s and -th versus zero in the third person plural present indicative with NP subject As regards the present indicative with adjacent third person plural NP subject, the suffixed verb, either with -s or -th, is the rule in the 16th-century texts except for a single occurrence of the unsuffixed variant in Stirlingshire in 1550. No instances of zero inflection have been recorded in North East, South Mid and Southern Scots in the 17th century. As regards other regions, the earliest examples of zero inflection in Northern letters date from 1660, in West Mid Scots from 1641, and in East Mid Scots from 1600. The suffixed verb represents the majority variant except in the early 18th-century letters representing Northern Scots. It is noteworthy that letters written by women contain no unsuffixed variants in the corpus in this context. With adjacent subjects containing singular NPs in co-ordination, the following present tense verb is also suffixed except in the early 18th-century letters. Another context to which the Northern Subject–Verb Rule applies is the present indicative in relative clauses with antecedents in the plural. In this context, the suffixed verb is the only variant in pre-1600 letters, and the unsuffixed verbs remain in the minority in the 17th century, and absent in Northern, North East, and South Mid Scots. The rare occurrences of zero inflection have been attested in East Mid Scots (Stirlingshire 1630, 1643, and Fife 1682), West Mid Scots (Lanarkshire 1659 [a female writer], 1692, and Ayrshire 1641) and in Southern Scots (Border counties 1600). The variant -th is rare in both of these contexts (only six occurrences). The th-variant is in a minority in the third person singular present indicative in 16thand 17th-century Scots. To stress its minority status, its spread across six regions is illustrated in Table 42.1 by mean frequencies: Table 42.1: Mean frequencies (/10,000) of the th-variant in the third person singular present indicative in six regions of Scotland in the CSC
1600–1649 1650–1715
Northern
North East
East Mid
West Mid
South Mid
Southern
0.4 3.0
1.2 8.2
5.4 3.5
4.2 6.7
7.5 0
4.3 0
The mean frequencies show that the th-variant was considerably less frequent in North East and Northern Scots in the first half of the 17th century, the highest having been attested in South Mid Scots. While the spread of the th-variant reached approximately the same level in post-1650 letters in North East, East Mid and West Mid Scots, it remains quite rare in Northern Scots and is absent in South Mid and Southern Scots.
3.3 Demonstrative pronouns In the Older Scots system of demonstrative pronouns, the plural of the proximal this is thir and the plural of the distal that is thai (see King 1997: 168–169; see also Beal
42 Early Modern English: Dialects
681
1997: 350–351). These and those occur almost exclusively in the 17th-century letters in the CSC (Meurman-Solin 2007a) data. In the 16th century, there is only one example in 1550 and another in 1596, both of them nominal uses. In a more detailed study of the Scottish English system of demonstratives, close attention should be paid to the fact that “the tendency to use these where Standard English would have those persists into the 20th century, at least in the phrase in these days, referring to the distant past” (Beal 1997: 350). Since thir is only used attributively, the following comparative account of the mean frequencies of thir, these, thai and those in four regions excludes nominal uses (for example, of the total of occurrences of these, approximately 60% are nominal uses). In examining the introduction of these and those and the decrease in thir and thai in East Mid Scots, we find developments of the following kind: 100 thir these thai those
80
60
40
20
0 1500–1549
1550 –1599
1600 –1649
1650 –1715
Figure 42.8: Percentages of plural demonstratives in East Mid Scots in the CSC
Figure 42.8 illustrates a remarkable change over time, with the diverging developments of thir and thai moving towards a lower frequency and these towards a higher. Figure 42.9 shows that thir, which is considerably more frequent in East Mid Scots, the basis of the so-called “Scottish Standard”, decreases in frequency most dramatically in this region during the first half of the 17th century. A time-lag can be seen in West Mid Scots in the same period: These brief illustrations in Section 3 show that space – here regions in Britain – and time – here the 16th and 17th centuries – can be conceptualized as autonomous variables in research based on sufficiently large corpora in which both variables belong to the core of rigorously defined compilation principles. In this purely quantitative approach, it has been possible to identify favored practices in a particular region at a particular time, as well as patterns in the chronology of spread across regions. However, as has been well documented in recent research, a thorough syntactic and semanticpragmatic analysis of these quantitative findings is required to understand fully the conditioning factors of variation and change. A further central dimension is the influence of contact between varieties.
682
V Early Modern English 25 East Mid Scots North East Scots West Mid Scots Northern Scots
20
15
10
5
0 1500–1549
1550 –1599
1600 –1649
1650 –1715
Figure 42.9: Mean frequencies (/10,000) of thir in four regions in the CSC (Meurman-Solin 2007a)
4 Further research Space and spatiality have been claimed to be largely untheorized variables in research on local and regional variation (Britain 2002: 603): Given the historical origins of variationism in traditional dialectology, and given the advances the discipline has made over the past decades in unpacking the initially rather crude attempts at understanding the social embedding of variation and change […], it is paradoxical that one of the social categories that has received least attention of all is space. Almost without exception, space has been treated as a blank stage on which sociolinguistic processes are enacted. It has been unexamined, untheorized, and its role in shaping and being shaped by variation and change untested.
Britain’s critical survey is highly relevant in assessing where the major advances and future challenges are in historical dialectology. The three dimensions of spatiality – Euclidean, social and perceived space (Britain 2002: 604) as well as models concerning knowledge about human geography in a particular region such as the description of the history of the Fens (Britain 2002: 604–606) – provide new insights into cross-disciplinary methodologies in the field. The approach to historical dialectology presented above (in Section 3) could be defined as data-driven and data-oriented. Methodologically, the claim here is that the study of language use in Euclidean space is complementary to that in social and perceived space in a highly relevant way, and, as shown by work at the Institute for Historical Dialectology in Edinburgh, also feasible with the assistance of sophisticated software applied to large quantities of data. In my view, combining the three conceptions of space into a complex merger of perspectives does not provide a methodologically useful tool. Instead, the three space types should be conceptualized as suggesting a sequence of perspectives, the first stage of data analysis taking place with reference to geographical coordinates, with time and space as primary variables, the data being then investigated from the perspective of social space, and, finally, from the perspective of perceived space.
42 Early Modern English: Dialects Even in this multi-perspective methodology, a dialectologist tends to marginalize the influence of discourse properties. However, since historical dialectology in the early periods of the English language exclusively draws on written data representing a wide range of different genres, space defined by discourse and text community type (see Section 2.2) is also a highly relevant conditioning factor. The present chapter claims that the most important advances in historical dialectology drawing on Early Modern English data can be found in research directly related to the creation of new digital databases. The creators of these new databases (Kyto¨, Grund, and Walker 2007; Kyto¨, Walter, and Grund 2007 in particular) base their compilation principles and practices on precisely defined and transparent criteria applicable to the assessment of diatopic representativeness. For example, to increase the relevance and validity of comparative descriptions over time and space, they restrict their work to one particular genre, trials, thus avoiding generalisation across genres. An equally relevant database can be extracted from corpora representing correspondence; however, it is usually necessary to shape a subcorpus, only including informants who are relatively homogeneous as regards variables such as education and geographical and social mobility. Overall, work towards a thorough description of regional variation in the Early Modern English period has only just started. Since the majority of texts now available in editions tend to represent the canon of literary and non-literary texts comparable to that used as source material for dictionaries, continued work towards making locally and regionally functional material on Early Modern English dialects available is required to improve the quality and quantity of research.
5 References Beal, Joan. 1997. Syntax and morphology. In: Jones (ed.), 335–377. Benson, Phil. 2001. Ethnocentrism and the English Dictionary. London/New York: Routledge. Britain, David. 2002. Space and spacial diffusion. In J. K. Chambers, Peter Trudgill, and Natalie Schilling-Estes (eds.), The Handbook of Language Variation and Change, 603–637. Malden, MA/Oxford: Blackwell. Caldwell, Sarah J. G. 1974. The Relative Pronoun in Early Scots. Helsinki: Socie´te´ Ne´ophilologique. Devitt, Amy J. 1989. Standardizing Written English. Diffusion in the Case of Scotland 1520–1659. Cambridge: Cambridge University Press. Dossena, Marina and Roger Lass (eds.). 2004. Methods and Data in English Historical Dialectology. Bern: Peter Lang. Dossena, Marina. 2005. Scotticisms in Grammar and Vocabulary. Edinburgh: John Donald Publishers. Go¨rlach, Manfred. 1996. And is it English? English World-Wide 17(2): 1–22. Go¨rlach, Manfred. 1999. Regional and social variation. In: Roger Lass (ed.), The Cambridge History of the English Language. Vol. III. 1476–1776, 459–538. Cambridge: Cambridge University Press. Grant, William and David Murison (eds.). 1931–76. The Scottish National Dictionary. Edinburgh: Scottish National Dictionary Association. Grund, Peter. 2007. From tongue to text: the transmission of the Salem witchcraft examination records. American Speech 82(2): 119–150. Grund, Peter, Merja Kyto¨ and Matti Rissanen. 2004. Editing the Salem Witchcraft Records: an exploration of a linguistic treasury. American Speech 79(2): 146–167. Huber, Magnus. 2007. The Old Bailey Proceedings, 1674–1834: Evaluating and annotating a corpus of 18th- and 19th-century spoken English. In: Anneli Meurman-Solin and Arja Nurmi (eds.), Annotating Variation and Change Vol. 1. http://www.helsinki.fi/varieng/journal/volumes/01/huber/
683
684
V Early Modern English Johnston, Paul. 1997. Older Scots phonology and its regional variation. In: Jones (ed.), 47–111. Jones, Charles (ed.). 1997. The Edinburgh History of the Scots Language. Edinburgh: Edinburgh University Press. Kastovsky, Dieter and Arthur Mettinger (eds.). 2000. The History of English in a Social Context. A Contribution to Historical Sociolinguistics. Berlin/New York: Mouton de Gruyter. King, Anne. 1997. The inflectional morphology of Older Scots. In: Jones (ed.), 156–181. Kyto¨, Merja and Terry Walker. 2003. The linguistic study of Early Modern English speech-related texts: How “bad” can “bad” data be? Journal of English Linguistics 31(3): 221–248. Kyto¨, Merja, Terry Walker, and Peter Grund. 2007. English witness depositions 1560–1760: An electronic text edition. ICAME Journal 31: 65–85. Kyto¨, Merja, Peter Grund, and Terry Walker. 2007. Regional variation and the language of English witness depositions 1560–1760: Constructing a “linguistic” edition in electronic form. In: Pa¨ivi Pahta, Irma Taavitsainen, Terttu Nevalainen, and Jukka Tyrkko¨ (eds.), Towards Multimedia in Corpus Studies Vol. 2. http://www.helsinki.fi/varieng/journal/volumes/02/kyto_et_al/ Laing, Margaret. 2004. Multidimensionality: time, space and stratigraphy in historical dialectology. In: Dossena and Lass (eds.), 49–96. Laing, Margaret and Roger Lass. 2006. Early Middle English dialectology: Problems and prospects. In: Ans van Kemenade and Bettelou Los (eds.), Handbook of the History of English, 417–451. Oxford: Blackwell. Laing, Margaret and Roger Lass. 2007. LAEME: A Linguistic Atlas of Early Middle English, 1150–1325. University of Edinburgh. http://www.lel.ed.ac.uk/ihd/laeme1/laeme1.html Lass, Roger. 2004. Ut custodiant litteras: Editions, corpora and witnesshood. In: Dossena and Lass (eds.), 21–48. McIntosh, Angus, Michael L. Samuels, and Michael Benskin. 1986. A Linguistic Atlas of Late Mediaeval English. 4 vols. Aberdeen: Aberdeen University Press. Meurman-Solin, Anneli. 1992. On the morphology of verbs in Middle Scots: Present and present perfect indicative. In Matti Rissanen, Ossi Ihalainen, Terttu Nevalainen, and Irma Taavitsainen (eds.), History of Englishes. New Methods and Interpretations in Historical Linguistics, 611–623. Berlin/New York: Mouton de Gruyter. Meurman-Solin, Anneli. 1993. Variation and Change in early Scottish Prose. Studies based on the Helsinki Corpus of Older Scots. Helsinki: Suomalainen Tiedeakatemia. Meurman-Solin, Anneli. 1997. Differentiation and standardization in early Scots. In: Jones (ed.), 3–23. Meurman-Solin, Anneli. 1999. Letters as a source of data for reconstructing early spoken Scots. In: Irma Taavitsainen, Gunnel Melchers, and Pa¨ivi Pahta (eds.), Writing in Nonstandard English, 305–322. Amsterdam/Philadelphia: John Benjamins. Meurman-Solin, Anneli. 2000a. On the conditioning of geographical and social distance in language variation and change in Renaissance Scots. In: Kastovsky and Mettinger (eds.), 227–255. Meurman-Solin, Anneli. 2000b. Geographical, socio-spatial and systemic distance in the spread of the relative who in Scots. In: Ricardo Bermu´dez-Otero, David Denison, Richard M. Hogg, and C. B. McCully (eds.), Generative Theory and Corpus Studies: A Dialogue from 10 ICEHL, 417– 438. Berlin/New York: Mouton de Gruyter. Meurman-Solin, Anneli. 2001. Women as informants in the reconstruction of geographically and socioculturally conditioned language variation and change in 16th and 17th century Scots. Scottish Language 20: 20–46. Meurman-Solin, Anneli. 2004. Data and methods in Scottish historical linguistics. In: Ermanno Barisone, Maria Luisa Maggioni, and Paola Tornaghi (eds.), The History of English and the Dynamics of Power, 25–42. Alessandria: Edizioni dell’Orso. Meurman-Solin, Anneli. 2005. Women’s Scots: Gender-based variation in Renaissance letters. In: Sally Mapstone (ed.), Older Scots Literature, 424–440. Edinburgh: John Donald Publishers. Meurman-Solin, Anneli. 2007a. Corpus of Scottish Correspondence (CSC). University of Helsinki. http://www.helsinki.fi/varieng/CoRD/corpora/CSC/; for manual, see: http://www.helsinki.fi/ varieng/csc/manual/.
43 Early Modern English: Language contact
685
Meurman-Solin, Anneli. 2007b. Annotating variational space over time. In: Anneli MeurmanSolin and Arja Nurmi (eds.), Annotating Variation and Change Vol. 1. http://www.helsinki.fi/ varieng/journal/volumes/01/meurman-solin/. Meurman-Solin, Anneli. 2007c. Relatives as sentence-level connectives. In: Ursula Lenker and Anneli Meurman-Solin (eds.), Connectives in the History of English, 255–287. Amsterdam/ Philadelphia: John Benjamins. Milroy, James. 1999. The consequences of standardisation in descriptive linguistics. In: Tony Bex and Richard J. Watts (eds.), Standard English: The Widening Debate, 16–39. London/New York: Routledge. Nevalainen, Terttu and Helena Raumolin-Brunberg. 1996. The Corpus of Early English Correspondence. In: Terttu Nevalainen and Helena Raumolin-Brunberg (eds.), Sociolinguistics and Language History. Studies based on the Corpus of Early English Correspondence, 39–54. Amsterdam/Atlanta, GA: Rodopi. Nevalainen, Terttu, Helena Raumolin-Brunberg, Jukka Kera¨nen, Minna Nevala, Arja Nurmi, and Minna Palander-Collin. 1998. Corpus of Early English Correspondence (CEEC). Department of English, University of Helsinki. http://www.helsinki.fi/varieng/CoRD/corpora/CEEC/index.html Nevalainen, Terttu and Helena Raumolin-Brunberg. 2000. The changing role of London on the linguistic map of Tudor and Stuart England. In: Kastovsky and Mettinger (eds.), 279–337. Nevalainen, Terttu and Helena Raumolin-Brunberg. 2003. Historical Sociolinguistics: Language Change in Tudor and Stuart England. London: Longman. Robinson, Mairi (ed.). 1985–96. Concise Scots Dictionary. Edinburgh: Chambers. Romaine, Suzanne. 1982. Socio-Historical Linguistics, Its Status and Methodology. Cambridge: Cambridge University Press. Williamson, Keith. 2004. On chronicity and space(s) in historical dialectology. In: Dossena and Lass (eds.), 97–136. Williamson, Keith. 2007. A Linguistic Atlas of Older Scots, Phase 1: 1380-1500 (LAOS). http:// www.lel.ed.ac.uk/ihd/laos1/laos1.html
Anneli Meurman-Solin, Helsinki (Finland)
43 Early Modern English: Language contact 1 2 3 4 5 6 7 8
Introduction Contact with Romance languages (the mixed-language business variety) Contact with Celtic languages Other foreign language contact within Britain Dialect contact within Britain English transported abroad Summary References
Abstract This chapter discusses contact between English and other languages, and also contact between different dialects within Britain. Such contact is likely to have led to grammatical changes, lexical borrowings and the rise of Standard English. Non-standard, predominantly Southern, English carried abroad by colonizers is also treated, as this is the period when English first became established in the New World and southern hemisphere. The Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 685–698
686
V Early Modern English main languages referred to in this chapter are Anglo-Norman, Medieval Latin, the Celtic languages spoken in Britain, Dutch, and French.
1 Introduction By 1475, English had already been greatly influenced by contact with other IndoEuropean languages, predominantly Old Norse and Anglo-Norman French. Spoken Anglo-Norman and Latin had influenced English well before the Early Modern period, but immediately prior to 1475 it seems to have been the written varieties of AngloNorman (AN) and Medieval Latin (MLt.), the languages of written record, which exerted an influence on business practice. Over the course of our period they were to be replaced by English, but their semantic and orthographic legacy continued. The first section considers the way in which Anglo-Norman and Medieval Latin were routinely mixed when keeping accounts.
2 Contact with Romance languages (the mixed-language business variety) Until around 1475 (earlier in the case of some archives, later in others), accounts and inventories were written in a mixed-language variety which combined Medieval Latin, Anglo-Norman, and Middle English. This might sound like a limited text-type, but keeping track of money in and money out is and was one of the fundamentals of business, and then as now was a major motivation for putting pen to paper. For example, here is a short extract from an account of money spent by the London Bridge House Estate in the financial year 1477–78 on repair. This is what business writing in the early part of the Early Modern period looked like, combining the three languages in an orderly fashion, (1): (1)
Item Cuidam harnden per CCxvj dies Johanni Boyle per CCxvj dies Thome Cowherd per CCviij dies henrico Milhale per CCxij dies & Johanni Reyne per lxvij dies fremasons operantes in hewyng & apparellyng lapides vocatis brigge assheler ac in factura & posicione illius nouo Groyne ad finem occidentalem Capelle sub tenemento in tenura Johannis Robgent / (London Metropolitan Archives, MS CLA/007/FN/02/003, London Bridge Wardens’ Annual Accounts volume 3, 1460–1484, folio 288, financial year 1477–8. Roman indicates expanded forms) ‘And to a certain Harnden for 216 days, John Boyle for 216 days, Thomas Cowherd for 208 days, Henry Milhale for 212 days and John Reyne for 67 days, freemasons, working in hewing and apparelling stone called bridge ashlar, and in making and positioning that new groin at the Western end of the chapel under the tenement in the tenure of John Robgent’
In this text-type the grammatical matrix is that of Medieval Latin, but nouns, adjectives, deverbal -ing forms, and proper names could optionally be realized in English, with both languages informed semantically by Anglo-Norman. In this Medieval Latin extract, there are the following non-Latin nouns and deverbal -ing forms: hewyng (OE+OE), apparellyng (AN+OE), brigge (OE), assheler (AN), groyne (?AN). This is
43 Early Modern English: Language contact not because the scribes did not know Latin; rather, a set of language-mixing rules had developed in this text-type, and plenty of documents written in this kind of mixed language still exist in archives around Britain. It was the supra-regional professional norm during the centuries preceding our period, and its effect on the development of written English (particularly with regard to spelling and lexis) is currently under investigation (see e.g. Wright 2000, 2001a, 2002a, 2002b, 2005; Trotter 2010; Ingham forthc.). Several other text-types routinely used language mixing, including medical, astronomical, mathematical, and other scientific texts; sermons and other religious writings; legal texts; administrative texts; literary prose, drama and verse (see Schendl 2000, 2002). So the first point is that Early Modern written English followed on directly from five hundred years’ worth of mixed-language usage, and carried over many of its Romanceinfluenced ways, including much technical vocabulary. At the beginning of the Early Modern period, writing was a still a predominantly trilingual affair, and monolingual writing in English was only just becoming the norm.
3 Contact with Celtic languages Within the British Isles, at the beginning of the Early Modern period, not everyone would have spoken English. The Celtic group of languages spoken in Britain – Welsh, Cornish, Irish Gaelic, Scottish Gaelic, Manx, perhaps Cumbrian – were mainly restricted to their various regions, but within those regions held sway. There has long been an acceptance that the Celtic languages did not exert much influence on English because there seem to be very few Celtic loan-words in English, or Celtic place-names in England (see Coates 2002: 72–75). This view has been countered by, amongst others, White (2003: 42), who points out that it is to be expected that a high-status peoples would be unlikely to borrow lexemes from a low-status substrate language, and that the relevant place-names refer to places that were only settled in late Anglo-Saxon times. Whether these Celtic languages caused any other kind of contact phenomena has been the subject of several recent papers, such as those in Tristram (ed.) (2000, 2003); and Filppula et al. (eds.) (2002). To consider those which fall within the remit of our period: Klemola (2000) suggests that the verb constraint known as the “Northern Subject Rule” might stem from substratum influence from the Brythonic group of Celtic languages, as it has parallels in Welsh, Cornish, and Breton. (The Northern Subject Rule operates in those Northern dialects where third-person plural present-tense markers are constrained by the adjacency of the pronoun they. An utterance from the Survey of English Dialects for Lancashire illustrates: they peel them and boils them – where the presence of they next to the verb peel blocks the suffix -s, which is otherwise realized on boils.) The Northern Subject Rule is first attested in Northern Middle English and Middle Scots texts, and is still current in some Scottish dialects. It can be found in Early Modern London English writing; Schendl (1996) analyzed third person plural present-tense indicative morphology in selected works of fifteen London authors and found it to be variably present. He analyzed selected Shakespearean texts, and found that “none of the c.160 instances of plural – (e)s in Shakespeare occurs in the pattern ‘they + adjacent plural present indicative verb’, though this construction is attested more than 300 times in Shakespeare’s works” (Schendl 1996: 150). With respect to the progressive, Mittendorf and Poppe (2000), Poppe (2003), and Filppula (2003) considered evidence for the possibility that the English progressive construction
687
688
V Early Modern English (to be + V-ing, as in Mary is singing) might stem from a Celtic substratum. Filppula (2003: 154–158) provides a summary of earlier scholars’ work, and opines that it is indeed likely to be the result of contact with Celtic languages. The progressive gained ground in the Early Modern period and is still widening its scope today. The spread and subsequent retrenchment of periphrastic do is one of the most well-known changes of our period. Klemola (2002) considers the rise of periphrastic do and its possible substratum in Brythonic Celtic (the Welsh, Cornish, Breton group). The earliest attestations of periphrastic do are found in affirmative declaratives in 13th-century South-western texts, and in negative declaratives and questions from the late 14th century (Denison 1993: 264–265). Periphrastic do in affirmative declaratives then began to decline during the latter half of the 16th century, standardizing to emphatic contexts only by about 1700 (see Warner, Chapter 47). Klemola (2002) suggests that the 19th-century geographical distribution of periphrastic do in Southwestern dialects, centering on Dorset, Wiltshire, Somerset, and Gloucestershire, indicates the probable Middle English heartland of this construction, and opines that Celtic influence is a likely contributory factor in the rise of periphrastic do. It should be highlighted that there is a distinction to be made between Celtic contact at point of settlement, resulting in the subjection of the Celts and their languages (murder, enslavement, exodus, and other types of cultural withdrawal, resulting in not much language contact phenomena – although with rather more place-name continuity than has hitherto been appreciated [Coates 2002: 73–75]), and prolonged Celtic contact with living speakers in the relevant parts of Britain culminating in our period, possibly resulting in such convergence features as the Northern Subject Rule, the progressive, and periphrastic do.
4 Other foreign language contact within Britain Turning now to speakers of languages from outside British territory: there is much evidence that other foreign languages were heard within Britain during the Early Modern period, but overall, and possibly with exceptions, it seems likely that their speakers accommodated to English rather than the other way around. Let us consider the evidence for their presence: two volumes by Goose and Luu (eds.)(2005) and Luu (2005) detail “alien” and “stranger” presence in Early Modern Britain. As a simplification, we can say that foreigners came to Britain either to trade, for economic betterment, or to escape persecution in their homelands. Of the latter, three groups stand out – Dutch, Flemish and Walloon refugees avoiding Spanish persecution in the southern Netherlands from the middle of the 16th century; the Moriscos, expelled from Spain in 1609; and the Huguenots who fled from France in the 1680s (Goose 2005: 1). Leaving aside London for the moment, All of the more significant provincial immigrant settlements are to be found in the south and east of England, in the towns of Norwich, Canterbury, Colchester, Sandwich, Maidstone, Southampton, Great Yarmouth, Dover, Thetford, King’s Lynn, Stamford, Halstead, Rye, Winchelsea, and later also on Canvey Island in Essex and in the fens of Lincolnshire and Cambridgeshire, at Sandtoft and Thorney respectively. Their foundations, with the exceptions of the fenland and Canvey Island communities, date from between 1561 and 1576 (Goose 2005: 17–18).
There were at least 16,000 aliens living in Canterbury, Colchester, London, Norwich, Sandwich, and Southampton by the 1570s (Luu 2005: 90), and “it is very likely that collectively the provincial immigrant communities outnumbered those in London in the
43 Early Modern English: Language contact late sixteenth century, and in total there may have been as many as 23,000–24,000 aliens in England by the 1590s” (Goose 2005: 17–18). The largest immigrant community outside London during our period was in Norwich. This began with the official importation of Flemish wool workers to boost the local economy in 1565. Next in size was Canterbury, with French-speaking Walloons manufacturing draperies and silks in 1575, reinforced by migrants from Calais in the 1640s, and the Huguenot immigration of the later 1600s. The Sandwich settlement of Flemings, also cloth-workers, dates from 1561. Numbers fluctuated over the decades, but suggested peak alien figures are Sandwich: 53% in the 1570s, Norwich: 40% in the 1580s, and Canterbury: 33% in the 1590s (Goose 2005: 18–21). These were the largest and longest-lasting immigrant communities outside London, where Dutch and French would have been routinely used. Trudgill (2001) has suggested that it is this process of language contact with Dutch and French that resulted in East Anglian leveling of the present-tense verb paradigm to zero, so that instead of thirdperson singular she walks, the traditional form in this area is she walk. Language contact is a plausible explanation for a morphological simplification of a system (as happened here), as it is well-known that speakers of mutually unintelligible languages who interact persistently in adulthood produce new systems that are morphologically simpler than the input systems – which is how pidgin languages are created. In fact, speakers in much of East Anglia had already been placed in a language contact situation by the advent of Viking Old Norse speakers in earlier centuries (see Poussa 1999; Nevalainen et al. 2001), so that the advent of the Dutch may be regarded as having exacerbated a language-contact process already set in motion. It is quite a challenge to disentangle processes of language contact and change initiated at the end of the first millennium from those introduced five hundred years later. By the mid-20th century the traditional English dialects had either leveled to -s or to zero (these being the only two options available by then, all other suffixes having been lost, with the exception of some vestigial usages of thou walkest), so all the dialects simplified with regard to present-tense verb morphology, with or without language contact. Turning now to London, Luu (2005: 87–140) provides data on immigrants in Elizabethan London, including whereabouts in the City they lived and where they originally came from. That it was cosmopolitan is not in doubt: Throughout all of the conflicts and controversies of the period London remained notably cosmopolitan, accommodating not only Dutch and French, but also small communities of Spanish, Italians, Germans and other nationalities, while there is little to indicate that – except for brief periods of national crisis or in relatively isolated incidents – foreign conflict ever translated into concerted action against the immigrants in our midst (Goose 2005: 9).
By 1483–84, the alien population in London had increased to 3,400, forming at least 6% of the population (Luu 2005: 89). The medieval Hansard and Italian communities were somewhat in decline in our period as trade shifted to Antwerp, with probably as few as seventy Italian and eighty Hanse merchants resident in London in the early 1500s (Luu 2005: 87). At this date there were possibly a further 1,800 artisan aliens living in London with their families, out of a total population of about 50,000 in 1500 (Goose 2005: 13). In 1550 a change to this pattern was introduced: a charter was granted to religious refugees to set up their own churches in London, resulting in waves of religious and political refugees, with perhaps 30,000 people entering England between 1567 and 1573. At its
689
690
V Early Modern English peak in 1553, London accommodated around 10,000 strangers out of an estimated total population of about 80,000 (Luu 2005: 88–91; Goose 2005: 13), forming roughly 12% of the population. Thereafter, numbers of immigrants declined in the capital, despite continued immigration, as more English people migrated to the city, and immigrants moved elsewhere in England or returned home. Estimates are down to about 10% by the 1570s (Luu 2005: 97). Even the late 17th-century Huguenot immigration probably did not significantly increase the proportion. With regard to languages spoken in the capital, in 1593 Luu estimates that Dutchspeakers constituted 55% of aliens in the capital, with French speakers constituting 34% and Italian speakers 3% (Luu 2005: 99). They settled overwhelmingly north of the river, concentrated largely on the edges of the city. There is little doubt that peoples of the same culture interacted together, but, of necessity, acquired English to a greater or lesser extent. To take one example, Wessell Webling, a brewer from Cleves, arrived in London around 1565 at the age of sixteen. He began work as a servant to his brother, a brewer in Southwark, and when his brother died in 1568 he continued in partnership with his brother’s widow. Webling not being as experienced in brewing as his brother, the business failed to make a profit and the partnership was dissolved. Webling then went to work with an English brewer who had married the widow of one of his brother’s alien colleagues, but this partnership, too, failed. By 1586, Webling had crossed the river and was living in the Steelyard. By 1593 he was running a large brewery employing more than thirty-four servants, and he eventually died a wealthy man in 1610, leaving a small bequest to the poor of his home town, Groten Recken in Westphalia, as well as larger sums to London friends and relatives (Luu 2005: 270, 276, 282). To return to 1576, when his fortunes were finally on the up but not yet established, we find him, along with a group of his countrymen, consorting with local women, (2): (2)
Wessell Weblinge beinge here called & chardged to haue offended wth mary Poyntell & to haue had thuse of her body He doth confesse yt that he is in falte ^therin^ and is content to geve of his benevolens to the poore xls Desyringe to be spared of his ponyshement And therefore the Courte is content to Dischardge hym for all offencs vntill this Daye/ (Bridewell Court Minute Book fo. 161v, 23 January 1576, London, Guildhall Library, MS Minutes of the Court of Governors of Bridewell and Bethlem: Microfilm Reels MS33011/1–2, 22 April 1559–6 May 1576)
Prostitution between foreign men and local women gave rise to language-contact opportunities in a weak-tie network context. Weak-tie networks are thought to be a conduit of language change, and the concept belongs to social network theory. Social network theory, initially developed in the field of sociology but adopted by biologists, geographers, and anthropologists as well as linguists, is a technique for accounting for the interactions between humans as they go about their lives in different social groupings – family, workplace, clique of friends, and so on. It was notably applied by Lesley Milroy (1987) when studying speech communities in Belfast, and has become a predominant paradigm in the field of historical sociolinguistics. Once a network of speakers has been identified the network is analyzed for density and multiplexity, that is, the frequency and kind of interaction that occurs between individuals. Ties between individuals are said to be strong if the individuals in a network interact
43 Early Modern English: Language contact frequently, repeatedly and in several social contexts, and to be strong ties if the individual members interact in this way with many other members of the same network. Ties are said to be weak if individuals interact infrequently in a single social context, and a loose-knit network with weak links is one where individuals interact with few others from that group on infrequent or single occasions. There are currently two predominant interpretations with regard to linguistic change and network ties. The first interpretation is that innovators of linguistic change are thought to have usually weak links to a network, but that strong ties facilitate actual adoption of innovations at a later stage. The second is that linguistic change is led by strong-tie innovators who sit at the hub of a web of high-frequency, multiplex strong and weak links. The first view has been developed by Milroy and Milroy (1985), the second by Labov (2001), but Raumolin-Brunberg (2006) has suggested that these linguists have been looking at different stages of the dissemination of linguistic change, with the Milroys concentrating on the initial stages, and Labov on the later developments. To conclude this section; there was systemic influence on English from Old Norse and Anglo-Norman (before the period considered here, but with continuing effect), written Anglo-Norman and Medieval Latin (in the early part of our period), and possibly the Celtic languages (with continuing effect in the Early Modern period); and dayto-day influence from speakers with other first languages (notably Dutch and French) in, particularly, London, Canterbury, Colchester, and the port towns of Norwich, Southampton, and Sandwich. Once immigrants had settled in Britain they embarked upon a lifelong process of English language acquisition and development.
5 Dialect contact within Britain Let us return to our immigrant, Wessell Webling. He maintained network ties with his countrymen for forty-five years, but also interacted with such English speakers as his employers (when he was younger) his workforce (when he was older), the younger members of his family, his local English friends, and community members. How can such lifelong influences on his English be reconstructed and analyzed, supposing him to have left sufficient data in the form of letters or other personal papers? Longitudinal studies over the lifespan are a recent development in the field of historical linguistics, the study of changes over an individual’s life having been made possible by the amassing of computer-searchable data. One such is Raumolin-Brunberg (2009), who looked at how Sir Walter Ralegh (from Devon), Philip Gawdy (from Norfolk) and John Chamberlain (Londoner) changed their grammar in adulthood with regard to the morphological variables my/thy vs mine/thine, -s vs -th, affirmative and negative do, and subject relativizer who, in a study of their personal letters. These changes seem to have occurred as a result of their shifting social networks as they moved around geographically and rose socially in Elizabethan society. In particular, Ralegh and Gawdy changed their grammar when they were middle-aged. Raumolin-Brunberg hypothesizes that either they just carried on increasing proportions of incoming variants acquired in childhood, or, as migrants to London, they accommodated to forms heard in their new environment, leading even to hypercorrection. She also notes that they, unlike Chamberlain, were socially ambitious, and that this ambition might be reflected in their adult language acquisition. Work detailed in Nevalainen and Raumolin-Brunberg (2003) charts the grammar of some more of these socially shifting migrants to London from other parts of
691
692
V Early Modern English Britain by analyzing personal letters held in the Corpus of Early English Correspondence for morphological variables. One of the great differences between the end of the Middle English period and the start of the Early Modern is the increase of visitors to the capital from both outside the capital and outside the country for reasons of trade (see Keene 2000; Wright 2001a) and social betterment. These included both migrants to the capital from other parts of Britain and the Continent, and people coming repeatedly on short-term visits and then returning home. I suggest that here we have a case of cause and effect: the cause being the increasing amount of visits to the capital from speakers outside it, and the effect being the rise of Standard English. Standard English bears the traces of a dialect-contact variety, from its simplified morphology to its mixed-dialect lexis. Whereas London English of the early Middle English period used predominantly regionally marked, Southern morphology, proto-Standard English (that is, English written for formal purposes from 1475 into the 1500s) came to adopt more and more linguistic features from the regional dialects, which features subsequently lost their regional marking. These include: – Southern present-tense indicative verb plural -th came to be replaced by Midland -en and -s. Subsequently, -th and -n were lost and zero became the Standard English marker, with -s still in use as a verb plural form in London English. – Southern present-tense indicative verb third person singular -th was replaced by Northern -s. – present participle markers changed from regionally marked -and(e), -end(e), -ind(e) to non-regional -ing. – Northern are took over as the plural third person present form of the verb to be and Southern be(n), beth were, eventually, lost. – South-western auxiliary do expanded in function. – adverbial suffixes reduced from Southern -liche to Northern -ly (or this change can be regarded as simply the loss of the regionally marked final syllable). – Northern pronouns they, them, their ousted hei, hem, here. Some of these developments can be regarded as the effect of dialect leveling, which is the process whereby interaction between speakers of different dialects leads to a kind of ‘consensus’ output, with features that are regionally marked lost, and features that are held in common retained. Thus, the regional plural indicative present-tense markers -th, -n, and -s were lost and zero adopted; non-regional present participle -ing took over from regionally marked -and(e), -end(e), -ind(e); and regional adverbial -liche went to non-regional -ly. In other cases, one regional form came to predominate in Standard English which, although originating outside the South-East, then lost its local marking. Third person singular -s, are, the th- pronouns, and do, all regional forms in late Middle English, no longer had regional connotations by the end of the Early Modern period (see the essays in Wright [ed.] (2000) for more on the processes of the standardization of English).
6 English transported abroad As English speakers began to colonize territories in the New World, the English language came into contact with indigenous languages, and also the languages of other
43 Early Modern English: Language contact colonizers. This situation led to the creation of New Englishes and English-lexicon creoles (Schneider 2007). As the bulk of this speaker contact lies outside our timeframe, in the Late Modern period (Hickey 2004a), I shall mention just the first Englishspeaking colony in America, the first in what is now Canada, the first Caribbean Englishes, and the first Southern Hemisphere English. The first English-speaking colony to survive successfully in North America was the Virginia Colony, with its township at Jamestown, founded by 104 Englishmen on 13 May 1607, with the first slaves thought to have arrived from Africa in 1619. The first young man to be sentenced to transportation to Virginia by the London Court of Bridewell was on 2 October 1607 (William Person, a dyer’s apprentice, found guilty of cozening (‘stealing’) his master’s goods and running away), and young people continued to be transported officially into the 1640s – and unofficially thereafter, due to the lucrative illegal practice of “spiriting” or kidnapping. Other London courts continued sentencing people to Virginia into the 1700s and to New England from 1643 (see, for example, Wright 2001b, 2003, 2004 and Wareing 2000, which analyzes data from the Middlesex Sessions Papers [1645–1718], recording the transporting of offenders to the New World from the Middlesex Courts). Not too much is known about early language contact with Algonkian speakers, although there is some record, and it is worth noting that native Americans went to London (Salmon 1992). Unfortunately, one at least seems to have come to grief there, (3): (3)
Kicko an Indian par Lord Maior kept. William Campion John Harding Humfrey Young Thomas Rich John Basse Beniamen Carter Thomas Clement Nicholas Moore William Pottes par Constable Cleworth Langborne vagrant boyes to be sent to Virginia (Bridewell Court Minute Book fo. 238, 26 August 1631, London, Guildhall Library, MS Minutes of the Court of Governors of Bridewell and Bethlem: Microfilm Reels MS33011/7, 1 March 1626–7 May 1634)
The main debates with regard to early American English have been about dialect influence on the development of American English from Ulster Scots (which lies substantially after our period) and about language contact with speakers of West African languages under conditions of slavery, putatively giving rise to present-day African American Vernacular English. This is a much-disputed topic, but as there is scant direct evidence from our period, it will not be pursued here (see Winford 1997, 1998 for a summary) (cf. Bailey, Volume 2, American English; Lanehardt, Volume 2, African American Vernacular English). The first English-speaking colony in what is now Canada was founded on the island of Newfoundland in 1583 (Clarke 2004). However, settlement patterns were seasonal, with residents returning to their homes in the West Country after one or two fishing seasons. The main language variety with which English came into contact in this locality was mostly Irish English (although there were some monolingual Irish Gaelic speakers) from 1675, but the bulk of the Irish immigration came considerably later (see Clarke 2004: 244–245). In the North Atlantic, Bermuda was colonized by English-speakers in 1609, with the first slaves on the islands by 1617. We hear of the first vagrants and petty criminals sentenced to Bermuda from the London Court of Bridewell in May 1619. The first Caribbean Islands to be colonized by English-speakers were St Kitts in 1623 and Barbados in 1626 (Baker 1998: 337–339; note that this is old-style dating – 1624 and 1627 are given as the founding dates in e.g. Hickey 2004b: 331). African slaves are known
693
694
V Early Modern English of on St Kitts from 1626, and young London vagrants were sentenced to St Kitts from June 1628 and to Barbados from January 1632. Initially, Barbados is thought to have had a hundred or more British settlers, thirty-two Amerindians from the Guianas, and in 1629 an influx of settlers from St Kitts (Baker 1998: 348) and in the 1640s from Ireland (Hickey 2004b: 331; for an overview of English-language contact in the Caribbean, see Hickey 2004b). Parkvall (1998: 64–5) gives information about the early colonization of St Kitts, with English speakers in contact with speakers of French and African languages during the 1600s. Baker (1998: 346–347) suggests that a language-contact “medium for inter-ethnic communication” would have come into being on St Kitts in the first few decades of settlement, although it should be noted that this view is contentious (cf. Hickey 2004b: 332). Baker’s argument is that previous commentators have assumed that slaves tried to speak the language of their masters, and that their English was effectively a learner variety with substrate interference, whilst the English-speakers simply addressed slaves in English. He posits another scenario: namely, all Kittitian speakers, freemen and slaves, would have tried to communicate with each other, thereby causing a medium for interethnic communication to come into being and eventually stabilize. This scenario holds good for all multilingual colonies consisting of adult monolingual speakers, and, indeed, explains why Surinam Creoles (for instance) have a predominantly English lexicon despite being only briefly under English rule (from 1651–67) and whence the majority of English settlers and Englishowned slaves had departed by 1675. If, as Baker suggests, an English-lexicon medium of interethnic communication had arisen before 1675, then there would have been no need to abandon it simply because the monolingual English-speakers had left and new Dutch administrators had arrived. It would still have functioned successfully as a code between slaves speaking non-mutually intelligible languages, and the Dutch administrators would have then spoken it, to a degree (Baker 1998: 348). Baker’s scenario can be supported by data from the oldest Southern Hemisphere English, that of the South Atlantic island of St Helena. From the late 1500s to 1673 St Helena was contested by the Portuguese, Dutch, French, and English. In 1673 the island was granted to the British East India Company in London to hold in perpetuity. The Company set up the island’s infrastructure, including keeping court records in the St Helena Consultations Books (British Library India Office G Factory Record Series MS G/32/2 St Helena Consultations 1676–1696), two copies of which survive. The East India Company and the free planters who settled on St Helena owned slaves from West Africa, Java, India, Madagascar, Sumatra, Borneo, and Malaya, amongst other places; and there is evidence in the St Helena Consultations that in the late 17th century such slaves spoke three or even four separate linguistic codes: a. first, the St Helena slaves are represented as speaking the same kind of non-Standard Southern English spoken by the free planters and the soldiers. This is the default language which the slaves and everyone else are recorded as speaking before the Court. It is possible that the Court Recorder standardized the slaves’ English and that it is presented as more competently spoken than it really was, but there is no evidence for this. b. second, some slaves reported that they could not understand others who spoke in Portuguese, a language used deliberately by some (rebelling) slaves so that other (non-rebelling) slaves would not understand. This may have been contemporary Portuguese, or a Portuguese-lexicon creole.
43 Early Modern English: Language contact c. third, in December 1695 the St Helena Consultations contain a small amount of a restructured or minimal-pidginized variety of English, mostly in negative constructions: “you noe savy”, “I noe tell you”, “me noe save spake English”, “you haue no good hart”, “Jack noe such foole”. There is very little, but as it is at such an early date it is important. It is compounded by the fact that some slaves are recorded as using elements of such pidgin to talk to each other as well as the Court, and the speakers who used it are also recorded as using English and hence are codeswitchers, possibly for social reasons. d. fourth, there is mention that some slaves spoke to others in their “country language”; that is, presumably, the language used in their country of origin (see Wright forthc.; Schreier and Wright 2010). It is also worth mentioning that a few lines containing similarly restructured English dating from a few years earlier, 1686, are extant in the Bodleian Library from a slave-trading station of the Gold Coast of West Africa (Huber 1999: 87–88). Huber (1999: 93) goes on to make the point that, wherever they were shipped to (and some went to St Helena), these slaves would have used such English (or restructured English) as they had gleaned for purposes of interethnic communication, so their variety cannot have been a fully expanded pidgin as such (expanded pidgins being unintelligible to speakers of the lexifier language). Finally, it should be emphasized that the variety of English inputting to all these New Englishes was, to a large extent, the regional variety of the people sentenced to go there from the London courts, or the poorer kind of planters and soldiers, and not Standard English. To conclude this section: there is little data on English language contact in the colonies during the Early Modern period, but as it was a significant, formative period for the resultant New Englishes and English-lexicon creoles, it deserves mention.
7 Summary Other types of Early Modern English language contact not dealt with in this chapter include the vast amount of word creation from Classical Latin and Greek roots that is to be found specifically in scientific writing (see, e.g., Norri 1992) and more generally in words of a formal register, and also borrowings due to specific activities, such as Italian in the field of music. The main thrust of this chapter has been to demonstrate that some Early Modern English speakers would have been, directly or indirectly, in a language-contact situation, whether because they came into contact with non-English speaking Celtic indigenes, because they came into contact with foreigners (with French and Dutch being the predominant languages involved), or because they came into contact with speakers from other dialect areas (with London being the largest city where this occurred). Simplification is one of the outcomes of language contact amongst adult speakers, and the processes of dialect contact are particularly significant in our period because one of the main differences between the Middle and the Early Modern periods is the tremendous amount of regional variation present in all the Middle English dialects as compared to the comparative lack of variation in Standard English. Had there been no contact between English speakers (especially those in London) and speakers of other languages, or contact between speakers of different English dialects, then English would not have standardized when and where it did.
695
696
V Early Modern English
8 References Baker, Philip. 1998. Investigating the origin and diffusion of shared features among the Atlantic English creoles. In: Baker and Bruyn (eds.), 315–364. Baker, Philip and Adrienne Bruyn (eds.). 1998. St Kitts and the Atlantic Creoles: The Texts of Samuel Augustus Mathews in Perspective. London: University of Westminster Press. Clarke, Sandra. 2004. The legacy of British and Irish English in Newfoundland. In: Hickey (ed.), 242–261. Coates, Richard. 2002. The significance of Celtic place-names in England. In: Filppula et al. (eds.), 47–85. Denison, David. 1993. English Historical Syntax: Verbal Constructions. London: Longman. Filppula, Markku. 2003. More on the English progressive and the Celtic connection. In: Tristram (ed.), 150–168. Filppula, Markku, Juhani Klemola, and Heli Pitka¨nen (eds.). 2002. The Celtic Roots of English. Joensuu: Joensuu University Press. Filppula, Markku, Juhani Klemola, and Heli Paulasto (eds.). 2008. English and Celtic in Contact. London: Routledge. Goose, Nigel. 2005. Immigrants in Tudor and Early Stuart England. In: Goose and Luu (eds.), 1–40. Goose, Nigel and Lien Luu (eds.). 2005. Immigrants in Tudor and Early Stuart England. Sussex Academic Press: Brighton. Hickey, Raymond. 2004a. Timeline for varieties of English. In: Hickey (ed.), 621–626. Hickey, Raymond. 2004b. English dialect input to the Caribbean. In: Hickey (ed.), 326–359. Hickey, Raymond (ed.). 2004. Legacies of Colonial English: Studies in Transported Dialects. Cambridge: Cambridge University Press. Huber, Magnus. 1999. Atlantic English creoles and the Lower Guinea Coast: A case against Afrogenesis. In: Magnus Huber and Mikael Parkvall (eds.), Spreading the Word: The Issue of Diffusion among the Atlantic Creoles, 81–110. London: University of Westminster Press. Ingham, Richard. forthc. Mixing languages on the manor. In: Judith Jefferson and Ad Putter (eds.), Medieval Multilingualism in Later Medieval Britain: Sources and Analysis. Turnhout: Brepols. Keene, Derek. 2000. Metropolitan values: Migration, mobility and cultural norms, London 1100– 1700. In: Wright (ed.), 93–116. Klemola, Juhani. 2000. The origins of the Northern Subject Rule: A case of early contact? In: Tristram (ed.), 329–346. Klemola, Juhani. 2002. Periphrastic DO: Dialectal distribution and origins. In: Filppula et al. (eds.), 199–212. Labov, William. 2001. Principles of Linguistic Change. Vol. II Social Factors. Oxford: Blackwell. Luu, Lien B. 2005. Immigrants and the Industries of London 1500–1700. Aldershot: Ashgate. Milroy, James and Lesley Milroy. 1985. Linguistic change, social network and speaker innovation. Journal of Linguistics 21: 339–384. Milroy, Lesley. 1987. Language and Social Networks. Oxford: Blackwell. Mittendorf, Ingo and Erich Poppe. 2000. Celtic contacts of the English progressive? In: Tristram (ed.), 117–145. Nevalainen, Terttu and Helena Raumolin-Brunberg. 2003. Historical Sociolinguistics: Language Change in Tudor and Stuart England. London: Longman Pearson Education. Nevalainen, Terttu, Helena Raumolin-Brunberg, and Peter Trudgill. 2001. Chapters in the social history of East Anglian English: The case of the third-person singular. In: Trudgill and Fisiak (eds.), 187–204. Norri, Juhani. 1992. Names of Sicknesses in English, 1400–1550: An Exploration of the Lexical Field. Helsinki: Suomalainen Tiedeakatemia. Orton, Harold, Eugen Dieth et al. 1962–71. Survey of English Dialects: The Basic Material. Leeds: E. J. Arnold. Parkvall, Mikael. 1998. A short note on the peopling of English St Kitts. In: Baker and Bruyn (eds.), 63–74.
43 Early Modern English: Language contact Poppe, Erich. 2003. Progress on the progressive? A report. In: Tristram (ed.), 65–84. Poussa, Patricia. 1999. Dickens as sociolinguist: Dialect in David Copperfield. In: Irma Taavitsainen, Gunnel Melchers, and Pa¨ivi Pahta (eds.), Writing in Nonstandard English, 27–44. Amsterdam/ Philadelphia: John Benjamins. Raumolin-Brunberg, Helena. 2006. Leaders of linguistic change in Early Modern England. In: Roberta Facchinetti and Matti Rissanen (eds.), Corpus-based Studies of Diachronic English, 115–134. Frankfurt am Main: Peter Lang. Raumolin-Brunberg, Helena. 2009. Lifespan changes in the language of three early modern gentlemen. In: Arja Nurmi, Minna Nevala, and Minna Palander-Collin (eds.), The Language of Daily Life in England (1400–1800), 165–196. Amsterdam/Philadelphia: John Benjamins. Salmon, Vivian. 1992. Thomas Harriot (1560–1621) and the origins of Amerindian Linguistics. Historiographia Linguistica 191: 25–56. Schendl, Herbert. 1996. The 3rd plural present indicative in Early Modern English – variation and linguistic contact. In: Derek Britton (ed.), English Historical Linguistics 1994. Papers from the 8th International Conference on English Historical Linguistics, 143–160. Amsterdam/Philadelphia: John Benjamins. Schendl, Herbert. 2000. Linguistic aspects of code-switching in medieval English texts. In: Trotter (ed.), 77–92. Schendl, Herbert. 2002. Mixed language texts as data and evidence in English historical linguistics. In: Donka Minkova and Robert Stockwell (eds.), Studies in the History of the English Language. A Millenial Perspective, 51–78. Berlin/New York: Mouton de Gruyter. Schneider, Edgar W. 2007. Postcolonial English: Varieties Around the World. Cambridge: Cambridge University Press. Schreier, Daniel and Laura Wright. 2010. Earliest St Helenian English in writing. In: Raymond Hickey (ed.), Varieties in Writing. The Written Word as Linguistic Evidence. Amsterdam/Philadelphia: John Benjamins. Tristram, Hildegard L. C. (ed.). 2000. Celtic Englishes II. Heidelberg: Universita¨tsverlag C. Winter. Tristram, Hildegard L. C. (ed.). 2003. The Celtic Englishes III. Heidelberg: Universita¨tsverlag C. Winter. Trotter, David. 2010. Bridging the gap: The (socio)linguistic evidence of some medieval English bridge accounts. In: Richard Ingham (ed.), The Anglo-Norman Language and its Context. York/Woodbridge: York Medieval Press in association with Boydell and Brewer. Trotter, David A. (ed.). 2000. Multilingualism in Later Medieval Britain. Woodbridge: D. S. Brewer. Trudgill, Peter. 2001. Third-person singular zero: African-American English, East Anglian dialects and Spanish persecution in the low countries. In: Trudgill and Fisiak (eds.), 179–186. Trudgill, Peter and Jacek Fisiak (eds.). 2001. East Anglian English. Woodbridge: D. S. Brewer. Wareing, John. 2000. The Regulation and Organisation of the Trade in Indentured Servants for the American Colonies in London, 1645–1718. Unpublished Ph.D dissertation, University of London. White, David L. 2003. Brittonic influence in the reductions of Middle English nominal morphology. In: Tristram (ed.), 29–45. Winford, Don. 1997. On the origins of African American Vernacular English – A creolist perspective. Part 1: Sociohistorical background. Diachronica 14(2): 305–344. Winford, Don. 1998. On the origins of African American Vernacular English – A creolist perspective. Part 2: Linguistic features. Diachronica 15(1): 99–154. Wright, Laura. 2000. Bills, accounts, inventories: Everyday trilingual activities in the business world of later medieval England. In: Trotter (ed.), 149–156. Wright, Laura. 2001a. The role of international and national trade in the standardisation of English. In: Isabel Moskowich-Spiegel Fandin˜o, Begon˜a Crespo Garcı´a, Emma Lezcano
697
698
V Early Modern English Gonza´lez, and Begon˜a Simal Gonza´lez (eds.), Re-interpretations of English. Essays on Language, Linguistics and Philology (I), 189–207. A Corun˜a: Universidade da Corun˜a. Wright, Laura. 2001b. Third person singular present-tense -s, -th and zero, 1575–1648. American Speech 76(3): 236–258. Wright, Laura. 2002a. Code-intermediate phenomena in medieval mixed-language business texts. Language Sciences 24: 471–489. Wright, Laura. 2002b. Standard English and the lexicon: Why so many different spellings? In: Mari C. Jones and Edith Esch (eds.), Language Change: The Interplay of Internal, External and Extra-linguistic Factors, 181–200. Berlin/New York: Mouton de Gruyter. Wright, Laura. 2003. Eight grammatical features of Southern United States speech present in Early Modern London prison narratives. In: Stephen J. Nagle and Sara L. Sanders (eds.), English in the Southern United States, 36–63. Cambridge: Cambridge University Press. Wright, Laura. 2004. The language of transported Londoners: Third-person-singular present-tense markers in depositions from Virginia and the Bermudas, 1607–1624. In: Hickey (ed.), 158–171. Wright, Laura. 2005. Medieval mixed-language business texts and the rise of Standard English. In: Janne Skaffari, Matti Peikola, Ruth Carroll, Risto Hiltunen, and Brita Wa˚rvik (eds.), Opening Windows on Texts and Discourses of the Past, 381–399. Amsterdam/Philadelphia: John Benjamins. Wright, Laura forthc. Black speakers on the island of St Helena, 1676–1715: Me noe save speake English. In: Michael D. Picone and Catherine Evans Davies (eds.), Proceedings of Language Varieties in the South III. Tuscaloosa: University of Alabama Press. Wright, Laura (ed.). 2000. The Development of Standard English 1300–1800: Theories, Descriptions, Conflicts. Cambridge: Cambridge University Press.
Laura Wright, Cambridge (UK)
44 Early Modern English: Standardization 1 2 3 4 5
Basic concepts Reduction of variation Elaboration of function Summary References
Abstract The chapter starts with a discussion of the concepts Early Modern English (EModE) and standardization. Early Modern English is established as the period between Middle English and Late Modern English; it covers the 16th and 17th centuries. Reduction of variation and elaboration of function are viewed as the processes which contributed most to EModE standardization. Reduction of variation is illustrated in Section 2, with examples from the levels of spelling, morphology, and syntax. Section 3 describes the conscious elaboration of the functions of the vernacular after it replaced Latin as the language of learned discourse. It started with the expansion of the vocabulary, which was necessary for the development of new text categories and changes in already existing ones. The last section summarizes the main issues of the chapter and places them in a broader historical perspective. Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 698–714
44 Early Modern English: Standardization
1 Basic concepts 1.1 Early Modern English Linguistic periods, like dialects, are abstractions. Their borderlines are fuzzy, and the criteria by which they are established are matters of dispute. There is not even agreement as to the types of criteria to be used; they can be language-internal or language-external or a combination of both (cf. Curzan, Volume 2, Chapter 79). Originally, historians of the English language worked with a binary model with the two periods Anglo-Saxon and English. At the end of the 19th century, it was replaced by a three-period model with the periods Old English, Middle English, and Modern English (Sweet 1891: 211). In this model, Early Modern English figures as a sub-period of Modern English, covering the time span 1500–1650. The most recent newcomer to the period model is Late Modern English. Although it already formed part of Sweet’s three-period model, where it figured as the sub-period from 1650 onwards, it is now conceived as a full-fledged period covering the 18th and 19th centuries. Its wide acceptance as a separate period is witnessed by a series of conferences, the first of which took place in Edinburgh in 2001. Late Modern English is perhaps too new a period to have caused discussions about its beginning and end. This is different for Early Modern English. Barber (1997) and Nevalainen (2006) agree on 1500 and 1700 as the beginning and the end of the EModE period without justifying these dates as more appropriate than others. But both scholars also mention other possible dates based on language-internal criteria. After reviewing earlier suggestions about the dating of Early Modern English, Go¨rlach (1991: 9–11), too, proposes 1500 and 1700 as its limits. The dates 1476–1776 in the title of volume III of the Cambridge History of the English Language mark a wider time span for Early Modern English with cultural and political events as points of demarcation. In the introduction to the volume, Lass (1999: 6) defends the extralinguistic event as the starting-point of a linguistic period by the repercussions it had on the text production process. The American Declaration of Independence as the end-point of the period is justified, because it initiated the development of English into a global language. Lass makes it quite clear that changes of the language system are related “in subtle and complex ways” to contemporary political, social, and cultural changes, but that neither can be considered the cause of the other. Therefore the extralinguistically defined end-point is not in conflict with his restriction of the linguistic period of EModE to the years 1500–1700 (Lass 1999: 9). Finally, there is a pragmatic reason for sticking to this periodization in the present chapter, too. Many of the studies on the development of individual features to be discussed in Section 3 are based on the Helsinki Corpus of English Texts (HC) (Rissanen et al. 1991), and in this invaluable research tool EModE is considered to cover the years 1500–1710.
1.2 Standardization We take it for granted that the general reference works on modern languages describe their standard varieties, i.e. the products of the process of standardization. Even when we consider only the written form of Present-day English (PDE), it is difficult to specify the criteria by which Standard English is defined. The Cambridge Grammar (Huddleston and Pullum 2002: 4) claims a “widespread agreement” about “what we are calling Standard English”; it relies on the variety used in broadcasting and in edited writing.
699
700
V Early Modern English The authors of the Comprehensive Grammar of the English Language (Quirk et al. 1985: 7) stress other properties; for them the standard variety is taught in the schools of the countries with English as a native language, and it is associated with educated users. Descriptions like these reflect the conviction that a standard language is an idea in the mind rather than a reality (Milroy and Milroy 1999: 19). Strictly speaking, this concept of a standard language is a paradox, because the two properties most intimately associated with standard languages are their lack of variation and their stability. Yet it is our experience that languages do change and that even standard languages allow a certain degree of variation. Milroy and Milroy (1999: 22–23) present a seven-step model for the description of the standardization process. It contains the stages selection, acceptance, diffusion, maintenance, elaboration of function, codification, and prescription. They point out that these stages need neither necessarily follow each other nor proceed with the same speed. The non-linear character of the standardization process is also stressed by Nevalainen and Tieken-Boon van Ostade (2006: 286) in their chapter “Standardisation” in the History of the English Language, and they attribute it “to the fact that it was not a consciously monitored development”. On this controversial point they agree with Hope (2000: 51–52), who claims that “[s]tandardisation […] looks more like a languageinternal phenomenon: something motivated and progressed below the conscious awareness of language users”. The overriding principle of EModE standardization was reduction of variation. The reduction process is partly due to the influence of public authorities, as in the case of the standardization of spelling; partly it is the outcome of subconscious choices. This is evident in all stages of the Milroyan model apart from prescription, and it will be illustrated on the levels of spelling, morphology, and syntax. The Milroys’ elaboration of function stage was a conscious process in Early Modern English in those fields where the vernacular replaced Latin. This will be illustrated from the field of the English vocabulary, which had to be enlarged before the language could take over new functions.
2 Reduction of variation Theoretically, variation can be reduced on the levels of spelling, grammar, and lexical semantics. The avoidance of polysemy was one of the goals of the natural philosophers, as the 17th-century representatives of the new experimental sciences were called. But they were realistic enough to see that this goal could not be achieved completely because of the infiniteness of the lexicon. Therefore, this aspect will be left out of consideration here. Variation was a pervasive feature of EModE spelling, morphology, and syntax, and most of the changes which took place in Early Modern English proceeded in the same direction, towards a reduction of variation. As it is not possible within the limits of this chapter to provide an exhaustive description of all standardization processes going on in Early Modern English, my aim will rather be to give an idea of the different paths reduction of variation can take.
2.1 Spelling EModE grammar writing followed the tradition of Latin grammar writing. Therefore, the first part of EModE grammars deals with the letters and their “voices”, i.e. their
44 Early Modern English: Standardization relation to sounds or phonemes. Yet before the first English grammars were produced, correct spelling was already an issue in England. Scragg (1974: 58–60) explains this ultimately as a consequence of the borrowing of many Latin words during the 16th century. Due to the prestige of Latin, the spelling of earlier loanwords, which had entered English via French and then been adapted to English phonology and word structure, was changed to make their impressive history more transparent. In some cases, both spellings still exist today, and each adopted a meaning of its own or got restricted to special contexts. Well-known examples are parson/person (< Lt. persona), poor/pauper (< Lt. pauper), frail/fragile (< Lt. fragilis). Very often, however, the more prestigious form replaced the older form completely, e.g. throne instead of trone, falcon instead of faucon, captive instead of caitif. The tendency to associate English words with Latin roots was so pronounced that it also affected words of native origin: the letter is an unetymological addition in scissors and scythe. They were wrongly associated with Lt. scindere ‘cut’, but the etymon of scythe is Old English sı¯ðe, and scissors derives from Late Lt. cisorium ‘a cutting instrument’. Association with a Latin “etymon” was not even necessary to promote spelling changes. On the analogy of dumb, where the final is etymologically motivated, the unetymological spellings crumb and thumb were introduced. It was only natural that the pronunciation of words such as throne, falcon, and captive was also changed to match the new prestigious spelling (spelling pronunciation). This change of pronunciation took some time so that different pronunciations were current for the same spelling. The coexistence of different spellings and different pronunciations for the same word gave rise to a heated debate about correct spelling. Whatever the principles were on which correct spelling should be based, it was clear from the outset that spelling must be stabilized. Two currents can be distinguished in the spelling reform debate; Nevalainen (2006: 32) refers to them as phonemic and logographic. The representatives of the former advocated the one phoneme–one letter principle, whereas those of the latter gave more room to tradition and etymology. John Hart (1569) was one of the fervent defenders of the one phoneme–one letter principle. He blamed the English spelling system because it used more letters in a word than it contained sounds. This is an attack against unetymological spellings of the type thumb, but also against the digraph in words like eight. He would not allow either that the same letter was used to represent two different phonemes. As a remedy for these shortcomings he proposed the following rules: Monophthongs are to be represented by a single vowel symbol, vowel length is to be indicated by a subscript dot; and always represent vowels, and consonants; represents the phoneme /k/; the letter is not used at all; for the representation of /tʃ/, /θ/ and /ð/, new symbols are invented. Although Hart’s system did not find much immediate approval, several early grammarians followed his line of argument; among them were William Bullokar (1580, 1586), Alexander Gil (1621), and Charles Butler (1634). In the preface to his grammar, Butler (1634: 7) complains about the lack of “a tru’ and constant writing”. He addresses two points, namely the variety of spellings for the same word, and the many-to-many relation between phoneme and letter. Both features lead to unnecessary difficulties for the learners of English. Butler (1634: 8) sees the reason for this deplorable situation in the inadequacy of the Latin alphabet, which has not enough letters for the representation of “all the single sounds of the English”. He blames contemporary spelling
701
702
V Early Modern English conventions for adopting letters from other alphabets and for using digraphs for the representation of one phoneme. He gives several examples to show that some letters do not evoke an association with the corresponding phonemes, e.g. the letters and representing the phonemes /j/ and /v/. Difficulties also arise from the coexistence of old and new spellings. Butler proposes a spelling reform based on three principles: new characters have to be introduced to compensate for the shortcomings of the Latin alphabet, some have to be given new and appropriate names, and the spelling has to mirror the generally accepted pronunciation of the words. He is sure that the introduction of this new spelling would save learners a lot of trouble and time. Butler’s grammar is written along the lines of this new spelling system. The importance Butler attaches to matters of spelling is evident from the layout of his grammar. It is divided into four chapters, of which the first two take up nearly half the grammar. They deal with letters and their distribution in syllables. All letters are used with three different fonts called Roman, Italic, and English, and they have upper case and lower case variants. Butler’s alphabet contains the 26 letters of the Latin alphabet, supplemented by the two digraphs and and eight special symbols. A final which is not pronounced is replaced by an apostrophe, e.g. plac’ for place. Butler himself was not too sure that his spelling would find many supporters. At the end of his note to the reader he expresses his fear that the power of custom might get the upper hand against all reason. Custom is indeed one of the key-words of the opposing party. In his First Part of the Elementarie, Richard Mulcaster (1582: 98) argued that a spelling reform was not necessary, because “[t]he vse & custom of our cuntrie, hath allredie chosen a kinde of penning, wherein she hath set down her relligion, hir lawes, hir priuat and publik dealings”. Since there is so much variation in pronunciation, he considers a phoneme-based spelling system undesirable. In his opinion consistency in spelling is much more important. As a schoolmaster he was primarily concerned with the teaching aspect of spelling. Therefore he included in his book an alphabetical list of about 8,000 words with their recommended spelling. Although this word-list became accepted as an authoritative reference work for spelling, the decisive role in the standardization process of spelling was played by the printers (Scragg 1974: 70; Nevalainen 2006: 36). One might have expected the regularizing influence of the printers to set in right after the introduction of the printing press by Caxton. But the first effect of the new technology was rather more than less variation, because Caxton had to employ foreign compositors, who were not familiar with English spelling conventions. Some progress was made when native English compositors were recruited. By the middle of the 16th century, quality printing houses followed the practices set up by the scriveners of the manuscript tradition. The variants which remained were not always due to the ignorance or the negligence of the printers, but rather a consequence of the exigencies of type-justification (Scragg 1974: 71–72). When there was too much space at the end of a line, the addition of a final was a handy remedy. Additionally the individual printing-houses did not follow the same spelling conventions. Nevertheless by 1650, the reduction of spelling variants had progressed so far that the system had nearly reached the present-day state. Two factors played an important role in this last stage of standardization: the influence of the spelling-books on the shaping of public taste and the effort of the printing-houses to cater for “the widest cross-section of the book-purchasing public” (Scragg 1974: 74).
44 Early Modern English: Standardization
2.2 Morphology Early Modern English as a transitional stage between earlier periods with more and later periods with fewer inflectional endings shows reduction of variation in all open word-classes, but also in its pronominal system. Although the morphology of substantives and adjectives was regularized as well, the most striking instances concern inflectional forms of verbs and especially the system of pronouns.
2.2.1 The third person singular indicative present tense In the 3P SG PRES IND variation was partly inherited from Middle English (ME), and partly it developed in Early Modern English. The endings -(e)th and -(e)s were used in Middle English, but they had a different regional distribution. The ending -(e)s prevailed in northern texts, -(e)th was typical of southern dialects. Stein (1987) distinguishes three stages in the EModE development from -(e)th to -(e)s in his study of literary texts. In stage A, which he documents with texts between 1525 and 1583, -(e)th is “the near-exclusive or 99% predominant ending” (Stein 1987: 407). Then follows stage B with variation between both endings. Their distribution depends on the stem-final phoneme of the verb and on matters of rhythm. The more modern ending is the rule, especially in negations; -(e)th is used after stem-final sibilants and when an additional syllable is required. The texts in Stein’s corpus which represent this stage date from 1572–1602. The first representative of stage C with a nearly exclusive ending -(e)s is Dekker’s The Seven Deadly Sinns of London (1606). Whereas Stein studied the -th/–s development only in literary texts, Bambas (1998) draws attention to the relevance of the text category for the spread of the -s ending. After analyzing 21 literary prose texts from the middle of the 16th to the middle of the 17th century, he comes to the conclusion that although the -s ending occurred only rarely in prose works before the last decade of the 16th century, it then held its ground firmly until the middle of the 17th century. He did not find any phonotactic or rhythmic constellations which favored one or the other ending. Variation between the two endings by the middle of the 17th century is also claimed by Nevalainen and Raumolin-Brunberg (2000: 238). It is interesting to note that Bambas supports his hypothesis of the survival of -th endings until well into the 17th century by Ben Jonson’s (1640) description of EModE verb morphology in his English Grammar and by his use of the -s ending in only 20% of all instances. All scholars who examined the development of the ending of the 3P SG PRES IND agree that at the end of the 17th century, the ending -(e)s was the norm. When -(e)th occurs in later texts, it is used for special stylistic effects.
2.2.2 The second person singular of the personal pronoun and of verbs One of the important changes towards standardization concerns verbs and pronouns at the same time. The ending -(e)st for the 2P SG PRES/PAST IND disappeared from the inflectional system of verbs, when the variation between the forms thou and you of the personal pronoun was given up in favor of you (cf. Nevalainen 2006: 100). Already in Middle English, the original plural form you had encroached upon the territory of the singular form thu, and in Early Modern English both forms were
703
704
V Early Modern English used side by side for some time (cf. Busse, Chapter 46, Section 2.5). Various models have been proposed to account for the distribution of the forms, most prominently the power and solidarity model (Brown and Gilman 1960, 1989). It predicts y–forms from +power speakers (i.e. speakers of the upper social ranks) in exchanges with their social equals and with –power speakers (i.e. speakers of the lower social ranks), but th-forms from –power speakers in exchanges with their social equals and with +power speakers. Yet this model does not satisfactorily explain many of the occurrences of either form (Hope 1993, 1994; Jucker 2000). In the present context it is important to note that whatever the motives may have been which led to the choice of one or the other form, the frequency of the th- forms (thou for subject case, thee for object case, thy/thine as possessive determiner and possessive pronoun) is steadily declining during the EModE period. Evidence from the HC and ARCHER, A Representative Corpus of Historical English Registers (Biber and Finegan 1990–93/2002/2007/2010), confirms this statement. Table 44.1: Tokens of thou in the EModE part of the HC (Rissanen et al. 1991) (per 10,000 words) Helsinki Corpus
Tokens of thou
1500–1570 1570–1640 1640–1710
24.45 17.59 12.16
In the HC (see Table 44.1) the figures of the sub-period 1640–1710 reflect the impact of the text category on the preservation of thou: 81 of the 208 tokens cluster in one text, namely Queen Elizabeth I’s translation of Boethius’s De consolatione philosophiae. Drama texts and dialogues in fiction texts as well as sermons provide another 61 examples. Plays and sermons are also the text categories which preserve thou best in the 18th century. The 18th century texts of ARCHER contain 54 tokens of thou, and c.60% of them occur in these two text categories. The parameter text category also proved relevant in Walker’s study of the distribution of thou and you in A Corpus of English Dialogues 1560–1760. She found that thou was best preserved in depositions (Walker 2007: 288, Figure 9.1). The influence of linguistic factors seems to depend on the choice of data. The hypothesis that thou is preferred with auxiliaries, you with lexical verbs (Barber 1997: 155, Lass 1999: 149) is supported by my data from the 18th-century ARCHER texts. With two exceptions the examples of thou are combined with auxiliaries. By contrast, Walker (2007: 293) found some preference for you in this environment in her corpus. When verbs or auxiliaries occur with thou, they preserve their inherited endings, i.e. -(e)st for lexical verbs in present and in past tense and -t in the forms art, wert, shalt, and wilt. When thou gets replaced by you, the unmarked forms of verbs and auxiliaries for the second person singular are used, and another instance of variation is lost through the spread of one form at the expense of the other.
2.2.3 The neutral possessive determiner My next example is more complex in that more than two variants are involved and more than one survived after the 17th century. The feature in question is the neuter
44 Early Modern English: Standardization possessive. The inherited form was his, which after the replacement of grammatical by natural gender of substantives was used in combination with animate and nonanimate head nouns. Its ME etymon had served as the genitive of the masculine and the neuter gender of the third person singular in the system of personal pronouns. The rise of the new form its at the end of the 16th century can be explained as an instance of gap-filling: the three-gender distinction of the personal pronoun was then paralleled by a three-gender distinction of the corresponding possessive determiners (Seppa¨nen 2000: 136). During the 16th century, his as a premodifier of nonanimate head-nouns had an infrequent competitor in the uninflected neuter personal pronoun it (Nevalainen and Raumolin-Brunberg 1994: 174). More frequent alternatives were the postnominal modifiers of it and thereof. The latter has the advantage that it can refer to animate and non-animate, to singular and plural substantives. The variant of the same was a minor alternative in official texts. The first grammarian to include the newcomer its into the paradigm of possessive determiners is Charles Butler (1634: 40). This is amazing for two reasons: first, it usually takes much longer before a new item finds its way into grammars; second, in the text of the grammar, Butler refers to non-animate substantives only by his (Moessner 2000: 409; Nevalainen and Raumolin-Brunberg 1994: 190). Quite ironically, the first grammarian who mentioned its in his grammar and consistently used it in the text of his grammar and in his other works is Guy Mie`ge, whose native tongue was French, and who published his grammar as late as (1688). Dryden’s disapproval of Ben Jonson’s use of his with reference to non-animate nouns indicates that its must have been the standard form in the second half of the 17th century (Moessner 2003: 46). This hypothesis is supported by quantitative evidence. Nevalainen and Raumolin-Brunberg (1994: 189) found a complementary distribution of his and it in the EModE part of the HC. If only these two variants are compared, his has a share of 100% in the first sub-period (1500–1570), and it is completely replaced by its in the third sub-period (1640–1710). In the second subperiod, his is still the predominant form, but there are already a few occurrences of its. The conclusion to be drawn in terms of standardization would be that by the middle of the 17th century, its had become the standard form. The situation is complicated by the existence of the additional variants of it and thereof in posthead position. In the first two sub-periods, when they competed with his, they had shares of between 30% and 37% if all variants are taken into account. Nevalainen and Raumolin-Brunberg (1994: 191) explain this high share with “the general tendency of placing items with animate reference before the head, and those with inanimate reference after the head”. When the gap was filled by its, the frequency of both post-modifiers dropped, and thereof became restricted to the genre of legal texts. Despite its frequency drop to 28% in the last sub-period, of it is the strongest competitor of its. At the beginning of the EModE period, the paradigm of possessive determiners contained the forms his, of it, thereof, and the minor alternatives it and of the same as variants of the third person singular neuter. At the end of the 16th century, its was added to the paradigm. By the middle of the 17th century, it had disappeared and his was replaced by its. From the second half of the 17th century, of the same and thereof became restricted to the legal register, its and of it remained the only competitors.
705
706
V Early Modern English
2.2.4 Comparison of adjectives Standardization followed a different path in the inflectional paradigm of adjectives. For the formation of the comparative and the superlative two patterns were available for the writers of EModE texts: the native English pattern with the endings -er and -est and the periphrastic pattern with more and most, which first appeared in the 13th century under the influence of Latin (Kyto¨ and Romaine 2000: 172). Most EModE grammarians treat these patterns as free variants, and those who give distribution rules do not agree about them (Dons 2004: 55–56). Modern accounts of EModE morphology add a third pattern, namely double comparison (Lass 1999: 158; Baugh and Cable 2002: 242; Barber 1997: 147; Go¨rlach 1991: 84), i.e. a combination of both patterns, as in the following example: (1)
The Duke of Milan And his more braver daughter could control thee (Shakespeare, The Tempest I.ii.439–440)
Dons’s (2004: 57–59) quantitative analysis of the HC shows that this strategy played only a minor role. When the form lesser is excluded, the percentage share of the double comparative is 2.12 in the sub-period 1500–1570 and drops to 0.68 in the sub-period 1570–1640. In the last sub-period double comparison was not used any longer. This is a clear case of reduction of variation. Changes in the distribution of the other two formation patterns are more complex. The inflectional pattern, which gained in frequency during the EModE period, continued this trend after the 17th century. Kyto¨ and Romaine (1997: 335) provide the following figures for inflectional comparatives: 55% (Late ME), 59% (EModE), 84% (PDE). Changes in the distribution of the two patterns seem to be an ongoing process, because in the period 1900–1950 the share of inflectional comparatives is only 69%. It should be borne in mind, however, that these figures do not tell us anything about the preferred strategy of comparison of individual adjectives. Criteria for the choice of one or the other strategy are similar in Early Modern English and in Present-day English: the number of syllables, the etymology of the adjective, the style level, and the text category.
2.3 Syntax The standardization processes covered here concern the structure of the verbal syntagm and the syntax of relative pronouns.
2.3.1 Multiple negation Multiple negation, a construction which was inherited from Middle English, was frequent in the 16th century (Rissanen 1999: 272). It involves the use of a negative particle together with negative determiners, pronouns, adverbs, or conjunctions. It occurred side by side with simple negation. Barber’s (1997: 199) claim that multiple negation was used for the purpose of emphasis does not hold for all cases. The examples below from the HC demonstrate that the same author could use both constructions for the same purpose, i.e. non-emphatic negation:
44 Early Modern English: Standardization a. But no Cristen man ys not suffered for to come ny it. (1517 Torkington, Ye Oldest Diarie of Englysshe Travell 30; HC) b. In thys Sepultur ys no Cristen man suffred to entre, (1517 Torkington, Ye Oldest Diarie of Englysshe Travell 36; HC) This is not to deny that there are instances of multiple negation which are adequately interpreted as emphatic, e.g.
(2)
(3)
I have one heart, one bosom, and one truth, And that no woman has, nor never none Shall mistress be of it, save I alone. (Shakespeare, Twelfth Night III.i.158–160)
These are the examples which Nevalainen (1998) considered in her study on the loss of multiple negation. Comparing the frequency of multiple and simple negation patterns for emphatic negation in the periods 1520–1550 and 1580–1610, she found that the share of multiple negation dropped from 40% to 12% from the first to the second period, and she concluded from these figures that multiple negation was “definitely on the way out around 1600” (Nevalainen 1998: 275). Since her data were coded for social rank and gender as well, the change can be described in more precise sociolinguistic terms. Men were leading the change, and they belonged to the ranks of the educated and upwardly mobile professionals. They were eager to imitate the new construction, which was quickly adopted by the members of the upper social ranks, and in their overzealousness they used it even more often than those they imitated. Nevalainen’s figures capture only changes in the frequency of the patterns of emphatic negation, and they do not allow quantitative statements about the development of multiple negation in the later part of the 17th century. More corpus research is needed, before we can make more precise statements about multiple negation than Barber’s (1997: 199): “It is rarely found in StE [Standard English] after the time of Shakespeare”.
2.3.2 The do-periphrasis The development of the do-periphrasis is one of the best researched patterns of English syntax (cf. Warner, Chapter 47). For nearly half a century the description provided by Ellega˚rd (1953) went unchallenged. Two reasons are responsible for the long and general acceptance of the results achieved in his study: they are based on a large corpus and obtained by the method of quantitative analysis. Very unusual for his time, Ellega˚rd employed statistical techniques for the interpretation of his figures. These are the stages which he established for the development of the do-periphrasis: Periphrastic do originated at the end of the 13th century and first occurred in poetic texts as a device for metrical and rhythmical purposes. At the end of the 15th century it became widely used in prose texts. Then it gained in popularity especially among the educated and literate circles. This development reached its peak in the middle of the 16th century. With the exception of affirmative questions, the frequency of periphrastic do dropped for several decades. From the last decade of the 16th century, the development of do in affirmative declarative sentences continued to drop, whereas in the other sentence types it rose steadily and by 1700 had become almost the rule. More recent investigations, adopting a sociolinguistic and variationist approach, provide evidence for an adjustment of Ellega˚rd’s
707
708
V Early Modern English timeline. On the basis of data from the Corpus of Early English Correspondence, Nurmi (1999) argues in favor of postponing the date of the decline of do in affirmative statements by about 40 years to the first decade of the 17th century. Since her data also document a decline of do in negative statements at the same time, she dates the beginning of the regularization process to the years after 1620, when the frequency of do in negative statements started to increase again. Very interestingly, she connects the frequency drop before to the accession of James I. His Scottish nobility introduced a new prestige variety with fewer occurrences of do. Whereas Nurmi shifted the completion of the standardization process already by about 20 years into the Late Modern English period, Tieken-Boon van Ostade (1987) in her study on the auxiliary do in the eighteenth century even suggests that it was an ongoing change in the 18th century. She analyzed informative prose, epistolary prose, and direct speech of 16 authors, and she found average shares of 21.42% of do-less negative sentences and of 9.79% of do-less questions in these text categories.
2.3.3 The syntax of relative pronouns The changes in the system of relative pronouns constitute a complex case of reduction of variation (cf. Johansson, Chapter 49). At the opening of the EModE period this system was realized by the forms who, which, the which, and that. All of them could refer to animate and to non-animate antecedents, and all of them could occur in restrictive and in non-restrictive attributive relative clauses. Already during the 16th century, the frequency of the which decreased, the form got restricted to prepositional syntagms and finally was dropped from the system: “Standardization appears to have worked here to the exclusion of one variant” (Raumolin-Brunberg 1996: 102). The frequency decline is convincingly illustrated by the figures from the HC: 72 occurrences in the first EModE sub-period (= 3.8/10,000 words), 21 occurrences in the second sub-period (1.1/10,000 words), in the third sub-period the which no longer occurs (Dons 2004: 77). This evidence is in line with the inclusion of the which in the inventory of relative pronouns in the grammars by Butler (1634) and Ben Jonson (1640). The remaining relative pronouns underwent standardization processes of a different kind. In the course of the 17th century the relative pronoun which became restricted to non-animate, the relative pronouns who and whom to animate antecedents, and the relativizer that to restrictive relative clauses (Rissanen 1999: 294).
3 Elaboration of function When at the beginning of the EModE period the vernacular took over in the domains which before were dominated by Latin and French, the inadequacy of English was felt in two fields, namely in matters of style and on the level of lexis. Lack of elegance and lack of copiousness were the shortcomings in the words of the commentators in the 16th century.
3.1 The need for new words Gaps in the vocabulary were a natural consequence of the quickly changing worldpicture: the earth as the center of the universe was replaced by the sun, the invention of the microscope and the telescope made objects visible which were invisible before, and the properties of blood circulation were discovered (Lass 1999: 1–2). The growth
44 Early Modern English: Standardization of specialized knowledge required new words for new things and concepts. Three strategies were available for the addition of new words to the English vocabulary: borrowing from other languages, the revival of obsolete native English words, and the exploitation of productive word-formation patterns. The growth of the vocabulary during the EModE period is truly remarkable (cf. Lancashire, Chapter 40). The calculations of new words for the period 1500–1700 range from 12,000 (Baugh and Cable 2002: 233) to 90,000 (Stockwell and Minkova 2001: 40–41). From the figures given in A Chronological English Dictionary (Finkenstaedt et al. 1970), it can be inferred that the 17th century contributed even more new words than the 16th century. Wermser (1976), who studied the expansion of the English vocabulary between 1450 and 1900, compared the share of neologisms in the common core of the vocabulary to that in specialized discourse. He found that the share of words belonging to specialized discourse increased throughout the EModE period and that the increase was particularly conspicuous in the natural sciences (Wermser 1976: 127–129). There is some disagreement on the main source of the new words. Baugh and Cable (2002: 230) state that “[b]y far the greater part of the additions to the English vocabulary in the period of the Renaissance was drawn from sources outside of English“. Based on Wermser’s figures for the individual sub-periods of his corpus, Nevalainen (1999: 351) concludes that “borrowing is by far the most common method of enriching the lexicon in Early Modern English”. These statements cannot be supported when the vocabulary of the natural sciences is analyzed. Robert Boyle, one of the most famous natural scientists and one of the founding members of the Royal Society, contributed as many as 446 neologisms. Only 73 (= 16.4%) are loan-words, but 315 (= 70.6%) belong to the category of affixations (Gotti 1996: 42–43). Henry Power, another scientist and member of the Royal Society, who contributed 109 lexemes to the EModE vocabulary, followed a similar strategy: 70.6% of his neologisms are affixations, and only 12.8% are loan-words. A comparison with the neologisms attested in William Barlow’s treatise Magneticall Aduertisements (1616) reveals that in the natural sciences, contrary to the common core vocabulary, affixations played a more important role than loan-words already at the beginning of the 17th century. Their share decreases between the first and the second half of the century, whereas the share of affixations increases (Moessner 2007: 250). A similar distribution of coining strategies is observed for the general vocabulary of the period by Barber (1997: 220) and by Stockwell and Minkova (2001: 41). The desire to enrich the English vocabulary must have led to an overuse of loanwords by some writers in the 16th century. Latin loan-words were an easy choice for those who were familiar with the classical languages, and they served well for filling gaps in the English vocabulary. The supporters of borrowing argued that the Romans used the same strategy and thus produced the copiousness for which Latin was so admired. The opposing party stressed the danger that loan-words were not only used in cases where an English word was not available, but also as a flourish of style to show off the writer’s learning. Additionally the meaning of the loan-words was not transparent, and this constituted a hindrance for the less educated part of the reading public. The different evaluation gave rise to the so-called “inkhorn controversy”, a public debate about the appropriateness of loan-words. Although the controversy had ebbed down in the 17th century, even a writer like Robert Boyle felt obliged to comment on the right use of loan-words. In his Proe¨mial Essay he writes:
709
710
V Early Modern English [A] writer may be allow’d to use Exotick Terms, especially when Custom has not only Denizon’d them, but brought them into request. […] [I]n Exotick Words, when Custom has once made them familiar and esteem’d, scrupulously to decline the use of them may be as well a fault, as needlessly to employ them (Hunter and Davis 1999: Vol. 2, 17).
Conscientious users of loan-words were aware of their obscurity, and they often added English paraphrases or more lengthy explanations. Barber (1997: 54) quotes Elyot with the paraphrases ‘gyue courage to others’ for animate, and ‘bringing vp of noble children’ for education. Paraphrasing of individual words as they occur in a text is an unsystematic device. Mulcaster (1582: 166) goes one step further and opts for the compilation of a dictionary: It were a thing verie praiseworthie in my opinion, and no lesse profitable then praise worthie, if som one well learned and as laborious a man, wold gather all the words which we vse in our English tung, whether naturall or incorporate, out of all professions, as well learned as not, into one dictionarie, and besides the right writing, …, wold open vnto vs therein, both their naturall force, and their proper vse.
Since he is above all concerned with spelling, the list of about 8,000 items, which he adds at the end of his book, contains only their recommended spelling, but no meanings. The first monolingual English dictionary with a sizeable number of entries is Edmund Coote’s The English Schoole-Maister (1596). It contains 1,368 lemmas in alphabetical order (Scha¨fer 1989: 42). Coote took over those items of Mulcaster’s list which would have caused problems for the less well educated readers, “some fewe of the hardest” as he describes them. For most of his entries he added glosses in a different font, e.g. accident befall, amorous full of loue. Some of his glosses are loan-words as well, e.g. epilogue conclusion, exquisite perfect. The entries which are not glossed are partly native words (e.g. almightie, bloud) and partly loan-words (e.g. circle, necessitie). All early lexicographers were primarily interested in these “hard words”, and their dictionaries specialized in these. The most important representatives of hard word dictionaries are Robert Cawdrey (A Table Alphabeticall 1604), John Bullokar (An English Expositor 1616), and Henry Cockeram (The English Dictionarie 1623). These were the first steps towards a standardization of the English vocabulary. The expansion of the vocabulary was a prerequisite for the development of new text categories and for changes in already existing ones. A good example of the former is the rise of the experimental essay (Moessner 2006), and crucial changes affected the category of medical writing (Taavitsainen 2004).
4 Summary Starting from the concept of EModE standardization as a process which is governed by the principle of reduction of variation and in which functional elaboration became necessary when Latin had yielded the field of learned discourse to the vernacular, several features of the English language between 1500 and 1700 were described which contributed to the development of modern Standard English. The treatment of standardization processes on the level of pronunciation was deliberately left aside. All through the EModE period there were at least two phoneme systems which existed side by side. Some speakers preferred the realization of the more
44 Early Modern English: Standardization progressive, others the realization of the more archaic one, and speakers could switch from one to the other. Furthermore, pronunciation is the level on which standardization has never been fully achieved: “When, however, we refer to ‘standard’ spoken English, we have to admit that a good deal of variety is tolerated in practice” (Milroy and Milroy 1999: 18). Attempts at reduction of variation were most successful on the plane of spelling, where by 1650 a spelling system was achieved which differs only little from the system still in use today. It was not my intention to describe all instances of variation reduction which took place on the grammatical level, both morphology and syntax, in Early Modern English. Rather I wanted to show the different ways in which standardization processes affected the grammatical system. Simple loss of one variant in favor of another one was illustrated by the replacement of the ending -(e)th by -(e)s in the third person singular indicative present of verbs, and by the complete replacement of multiple negation by simple negation. A similarly straightforward, but different type was illustrated by the development of periphrastic do, where the originally free variants of verb forms with and without do developed into variants in complementary distribution. The case of the replacement of thou by you showed that standardization processes can also proceed pairwise; with the loss of thou the ending -(e)st of the second person singular indicative of verbs disappeared, too. More complex processes are involved in the changes affecting the comparison of adjectives and the syntax of relative clauses. In both cases one variant got lost, namely double comparatives and the which respectively, and the other variants became restricted to special environments: the inflectional comparatives and superlatives to monosyllabic adjectives and the analytic forms to polysyllabic adjectives, who to animate and which to non-animate antecedents. The most complex constellation concerned the possessive determiner its. Here a completely new form entered the paradigm, where it competed with his, it, of it, and thereof. Although its is first attested only in 1598, at the end of the EModE period it had ousted his and it completely and pushed thereof into the niche of the legal register. When looking at the individual processes, it became evident that they proceeded at different pace and that some were completed at the end of the EModE period (e.g. the loss of multiple negation, double comparatives), whereas others have to be described as change in progress (e.g. periphrastic do, the distribution of synthetic and analytic comparative formation). These are the standardization processes which need to be followed up in the analysis of Late Modern English. Functional extension required first of all an expansion of the English vocabulary. Lack of copiousness was indeed one of the complaints of the early commentators. Translators and the writers of texts about specialized topics felt the need for new words especially badly. A line of development was traced from statements about the necessity of authoritative word-lists to the compilation of the first monolingual English dictionaries and the rise of the experimental essay as a new genre and the conceptual and linguistic changes in the genre of medical texts. The textual representatives of these genres had reached characteristic profiles at the end of the EModE period, but they were subject to later changes. Standardization processes helped to shape these profiles, but they did not produce stable text categories.
711
712
V Early Modern English
5 References Bambas, Rudolf. 1998. Verb forms in -s and -th in Early Modern English prose. In: Mats Ryde´n, Ingrid Tieken-Boon van Ostade, and Merja Kyto¨ (eds.), A Reader in Early Modern English, 65–71. Frankfurt/Main: Peter Lang. Barber, Charles. 1997. Early Modern English. 2nd edn. Edinburgh: Edinburgh University Press. Biber, Douglas and Edward Finegan. 1990–1993/2002/2007/2010. A Representative Corpus of Historical English Registers (ARCHER). Version 3.1 http://www.llc.manchester.ac.uk/research/ projects/archer/ Baugh, Albert C. and Thomas Cable. 2002. A History of the English Language. 5th edn. London: Routledge. Brown, Roger and Albert Gilman. 1989. Politeness theory and Shakespeare’s four major tragedies. Language in Society 18: 159–212. Brown, Roger, and Albert Gilman. 1960. The pronouns of power and solidarity. In: Thomas A. Sebeok (ed.), Style in Language, 253–276. Cambridge, MA: MIT Press. Bullokar, John. 1616. An English expositor. (English Linguistics 1500–1800, ed. by R. C. Alston, 11.) Menston: The Scolar Press, 1967. Bullokar, William. 1580, 1586. Book at Large (1580) and Bref Grammar for English (1586). Facsimile reproductions with an introduction by Diane Bornstein. New York: Scholar’s Facsimiles & Reprints, 1977. Butler, Charles. 1634. The English Grammar, or The Institution of Letters, Syllables, and Woords in the English Tung. Oxford: William Turner. [Albert Eichler (ed.), Charles Butler’s English Grammar. Halle a. S.: Max Niemeyer, 1910.] Cawdrey, Robert. 1604. A Table Alphabeticall. (English Experience, its Record in Early Printed Books Published in Facsimile, 226.) New York: Da Capo Press, 1970. Cockeram, Henry. 1623. The English Dictionarie. (English Linguistics 1500–1800, ed. by R. C. Alston, 124.) Menston: The Scolar Press, 1968. Coote, Edmund. 1596. The English Schoole-Maister. (English Linguistics 1500–1800, ed. by R. C. Alston, 98.) Menston: The Scolar Press, 1968. Dalton-Puffer, Christiane and Nikolaus Ritt (eds.). 2000. Words: Structure, Meaning, Function. A Festschrift for Dieter Kastovsky. Berlin/New York: Mouton de Gruyter. Dons, Ute. 2004. Descriptive Adequacy of Early Modern English Grammars. Berlin/New York: Mouton de Gruyter. Ellega˚rd, Alvar. 1953. The Auxiliary do. The Establishment and Regulation of its Use in English. Stockholm: Almqvist & Wiksell. Finkenstaedt, Thomas, Ernst Leisi, and Dieter Wolff. 1970. A Chronological English Dictionary. Heidelberg: Winter. Gil, Alexander. 1621. Logonomia Anglica. 2nd edn. (English Linguistics 1500–1800, ed. by R. C. Alston, 68.) Menston: The Scolar Press, 1968. Go¨rlach, Manfred. 1991. Introduction to Early Modern English. Cambridge: Cambridge University Press. Gotti, Maurizio. 1996. Robert Boyle and the Language of Science. Milan: Guerini. Hart, John. 1569. An Orthographie. (English Linguistics 1500–1800, ed. by R. C. Alston, 209.) Menston: The Scolar Press, 1969. Hope, Jonathan. 1993. Second person singular pronouns in records of Early Modern “spoken” English. Neuphilologische Mitteilungen 94: 83–100. Hope, Jonathan. 1994. The use of thou and you in Early Modern spoken English: Evidence from depositions in the Durham ecclesiastical court records. In: Dieter Kastovsky (ed.), Studies in Early Modern English, 141–151. Berlin/New York: Mouton de Gruyter. Hope, Jonathan. 2000. Rats, bats, sparrows and dogs: Biology, linguistics and the nature of Standard English. In: Wright (ed.), 49–56.
44 Early Modern English: Standardization Huddleston, Rodney and Geoffrey K. Pullum. 2002. The Cambridge Grammar of the English Language. Cambridge: Cambridge University Press. Hunter, Michael and Edward B. Davis (eds.). 1999–2000. The Works of Robert Boyle. 14 vols. London: Pickering and Chatto. Jonson, Ben. 1640. The English Grammar (from the works). (English Linguistics 1500–1800, ed. by R. C. Alston, 349.) Menston: The Scolar Press, 1972. Jucker, Andreas H. 2000. Thou in the history of English: A case for historical semantics or pragmatics? In: Dalton-Puffer and Ritt (eds.), 153–163. Kyto¨, Merja and Suzanne Romaine. 1997. Competing forms of adjective comparison in Modern English: What could be more quicker and easier and more effective? In: Terttu Nevalainen and Leena Kahlas-Tarkka (eds.), To Explain the Present. Studies in the Changing English Language in Honour of Matti Rissanen, 329–352. Helsinki: Socie´te´ Ne´ophilologique. Kyto¨, Merja and Suzanne Romaine. 2000. Adjective comparison and standardisation processes in American and British English from 1620 to the present. In: Wright (ed.), 171–194. Lass, Roger. 1999. Introduction. In: Lass (ed.), 1–12. Lass, Roger. 1999. Phonology and morphology. In: Lass (ed.), 56–186. Lass, Roger (ed.). 1999. The Cambridge History of the English Language. Vol. III. 1476–1776. Cambridge: Cambridge University Press. Mie`ge, Guy. 1688. The English Grammar. (English Linguistics 1500–1800, ed. by R. C. Alston, 152) Menston: The Scolar Press, 1969. Milroy, James and Lesley Milroy. 1999. Authority in Language. Investigating Language Prescription and Standardisation. 2nd edn. London/New York: Routledge. Moessner, Lilo. 2000. Grammatical description and language use in the seventeenth century. In: Ricardo Bermu´dez-Otero, David Denison, Richard M. Hogg, and C. B. McCully (eds.), Generative Theory and Corpus Studies. A Dialogue from 10 ICEHL, 395–416. Berlin/New York: Mouton de Gruyter. Moessner, Lilo. 2003. Diachronic English Linguistics. An Introduction. Tu¨bingen: Gunter Narr. Moessner, Lilo. 2006. The birth of the experimental essay. In: Vijay K. Bhatia and Maurizio Gotti (eds.), Explorations in Specialized Genres, 59–77. Bern: Peter Lang. Moessner, Lilo. 2007. The vocabulary of Early Modern English scientific texts. In: Ute Smit, Stefan Dollinger, Julia Hu¨ttner, Gunther Kaltenbo¨ck, and Ursula Lutzky (eds.), Tracing English through Time. Explorations in Language Variation, 235–252. Wien: Braunmu¨ller. Mulcaster, Richard. 1582. The First Part of the Elementarie which entreateth chefelie of the right writing of our English tung. London: Thomas Vautroullier. Nevalainen, Terttu. 1998. Social mobility and the decline of multiple negation in Early Modern English. In: Jacek Fisiak and Marcin Krygier (eds.), Advances in English Historical Linguistics (1996), 263–291. Berlin/New York: Mouton de Gruyter. Nevalainen, Terttu. 1999. Early Modern English lexis and semantics. In: Lass (ed.), 332–456. Nevalainen, Terttu. 2006. An Introduction to Early Modern English. Edinburgh: Edinburgh University Press. Nevalainen, Terttu and Helena Raumolin-Brunberg. 1994. Its strength and the beauty of it: The standardization of the third person neuter possessive in Early Modern English. In: Dieter Stein and Ingrid Tieken-Boon van Ostade (eds.), Towards a Standard English 1600 – 1800, 171–216. Berlin/New York: Mouton de Gruyter. Nevalainen, Terttu and Helena Raumolin-Brunberg. 2000. The third-person singular -(e)s and -(e)th revisited: The morphophonemic hypothesis. In: Dalton-Puffer and Ritt (eds.), 235–248. Nevalainen, Terttu and Ingrid Tieken-Boon van Ostade. 2006. Standardisation. In: Richard Hogg and David Denison (eds.), A History of the English Language, 271–311. Cambridge: Cambridge University Press. Nurmi, Arja. 1999. A Social History of Periphrastic DO. Helsinki: Socie´te´ Ne´ophilologique. Quirk, Randolph, Sidney Greenbaum, Geoffrey Leech, and Jan Svartvik. 1985. A Comprehensive Grammar of the English Language. London/New York: Longman.
713
714
V Early Modern English Raumolin-Brunberg, Helena. 1996. Apparent time. In: Terttu Nevalainen and Helena RaumolinBrunberg (eds.), Sociolinguistics and Language History. Studies based on the Corpus of Early English Correspondence, 93–109. Amsterdam/Atlanta: Rodopi. Rissanen, Matti, Merja Kyto¨, Leena Kahlas-Tarkka, Matti Kilpio¨, Saara Nevanlinna, Irma Taavitsainen, Terttu Nevalainen, and Helena Raumolin-Brunberg. 1991. The Helsinki Corpus of English Texts. In: ICAME Collection of English Language Corpora (CD-ROM), 2nd edn., Knut Hofland, Anne Lindebjerg, and Jørn Thunestvedt (eds.), The HIT Centre, University of Bergen, Norway. For manual, see http://khnt.hit.uib.no/icame/manuals/hc/index.htm Rissanen, Matti. 1999. Syntax. In: Lass (ed.), 187–331. Scha¨fer, Ju¨rgen. 1989. Early Modern English Lexicography. Vol. I. A Survey of Monolingual Printed Glossaries and Dictionaries 1475–1640. Oxford: Clarendon Press. Scragg, Donald George. 1974. A History of English Spelling. Manchester: Manchester University Press. Seppa¨nen, Aimo. 2000. The genitive/possessive pronoun its. Studia Neophilologica 72: 121–141. Shakespeare, William. 1997. The Complete Works. 2nd edn. G. Blakemore Evans (ed.). Boston/ New York: Houghton Mifflin. Stein, Dieter. 1987. At the crossroads of philology, linguistics and semiotics: Notes on the replacement of th by s in the third person singular in English. English Studies 68(5): 406–432. Stockwell, Robert and Donka Minkova. 2001. English Words: History and Structure. Cambridge: Cambridge University Press. Sweet, Henry. 1891. A New English Grammar Logical and Historical. Part I. Oxford: Clarendon Press. Taavitsainen, Irma. 2004. Transferring classical discourse convention into the vernacular. In: Irma Taavitsainen and Pa¨ivi Pahta (eds.), Medical and Scientific Writing in Late Medieval English, 37–72. Cambridge: Cambridge University Press. Tieken-Boon van Ostade, Ingrid. 1987. The Auxiliary DO in Eighteenth-century English. A Sociohistorical-linguistic Approach. Dordrecht: Foris Publications. Walker, Terry. 2007. Thou and You in Early Modern English Dialogues. Amsterdam/Philadelphia: John Benjamins. Wermser, Richard. 1976. Statistische Studien zur Entwicklung des englischen Wortschatzes. Bern: Francke Verlag. Wright, Laura (ed.). 2000. The Development of Standard English, 1300–1800: Theories, Descriptions, Conflicts. Cambridge: Cambridge University Press.
Lilo Moessner, Aachen (Germany)
45 Early Modern English: Sociolinguistics 1 2 3 4 5
Sociolinguistics and language history Early Modern England 1500–1700 The sociolinguistic patterns of language change Summary References
Abstract This chapter shows that there was considerable sociolinguistic variation in Early Modern England. Despite the problems of accessing the lowest social ranks and women due to Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 714–731
45 Early Modern English: Sociolinguistics widespread illiteracy, studies of letter and dialogue data testify to the significance of gender, social rank, region, and register in the diffusion of linguistic change. The chapter focuses on morphosyntactic changes, such as the replacement of the subject pronoun ye by you, the adoption of the third-person singular suffix -s instead of -th and the loss of multiple negation. As today, women seemed to lead changes from below but, unlike today, changes from above were led by men. Social stratification also proved significant, and the capital region, London and the Court, formed the centre from which changes spread elsewhere in the country. The chapter also shows that some individuals changed their language across their lifespans, and weak-tie networks apparently promoted the diffusion of change.
1 Sociolinguistics and language history Basically, sociolinguistic variation has been approached from two complementary angles, those of the speaker and of the situation of language use. The linguistic choices individual language users make have been observed to correlate with their social backgrounds, i.e. gender, social status, education, ethnic group, etc. Moreover, linguistic choices have also been found to be influenced by the situation of language use and, in particular, the relationship between the interlocutors. Beside these topics, sociolinguistic research has covered other areas of study, such as standardization and prescription, language attitudes, social networks, patterns of communication, and multilingualism, to mention just a few. Most of what we know of sociolinguistic variation stems from studies of present-day languages, but there is no reason to believe that similar phenomena would not have existed in the past. Historical and cultural research has given ample evidence of the existence of social and gender stratification in past societies, for instance, and we can assume that language variation in the past was constrained by social factors, although they were not necessarily the same as in today’s societies. Present-day sociolinguistics and historical linguistics share one important objective: both strive to account for linguistic change. In recent years, historical linguists have begun to apply sociolinguistic methods to their research. The forerunner in this field was Romaine (1982), who used genre variation to trace changes in the relative pronoun system in Middle Scots. A large number of studies using similar methods have followed suit, and research has since also been extended beyond texts to individual language users in the study of one informal genre such as personal correspondence (as in Nevalainen and Raumolin-Brunberg 2003). Although sociolinguists, both present-day and historical, take linguistic change as their central topic, there is no reason to claim that all sociolinguistic variation should involve change. However, because stable variation has only received limited attention in historical studies (see, e.g., Raumolin-Brunberg 2002; Laitinen 2007: 212), it seems appropriate here to concentrate on phenomena that have been uncovered in the studies of early modern linguistic changes. While sociolinguists studying present-day languages can use informal spoken language as their data, historical linguists focusing on developments before the innovation of audio recording have only written language for their use. As regards Early Modern English, there is a great deal of written data available from various genres, some of which are rather informal, representing language that, at least to some extent, resembles the spoken
715
716
V Early Modern English idiom of the time. These genres include personal letters and dialogues, which form the main sources that have been used for research in historical sociolinguistics. Letters represent the written mode of language, and illiteracy sets its limits to their use as research data. On the other hand, they are good data for sociolinguistic research, as present-day studies (e.g. Biber 1988) show that personal letters in many ways resemble such spoken genres as spontaneous speech and interviews. It is also important to notice that they comprise genuine interaction (for further information on letters as a genre, see Nurmi and Palander-Collin 2008). In addition, the writers and recipients of letters can be traced and this information used for sociolinguistic analysis. The historical study of dialogue texts such as trial proceedings, witness depositions, and drama is not unproblematic either because of the inaccurate ways spoken language has been recorded in writing or the poor ability of many playwrights to imitate speech. However, as Kyto¨ and Walker (2003: 241) argue, there are speech-related texts which may be fairly reliable records of spoken interaction of the past and hence relevant material for historical sociolinguistics. The development of electronic text corpora and retrieval programs has been instrumental for historical sociolinguistic research. Most of the material discussed in this article goes back to the electronic Corpus of Early English Correspondence (CEEC) (Nevalainen et al. 1998), which was compiled at the University of Helsinki for research in historical sociolinguistics in 1993−97. The CEEC, covering the period c.1410−1681, contains over 6000 letters, c.2.7 million words, from 778 informants (for further information, see Raumolin-Brunberg and Nevalainen 2007). Material from the Corpus of Early English Correspondence Supplement (CEECSU) (Kaislaniemi et al. 2000) covering the period 1402–1663 (c.900 letters and 0.44 million words) was also used. A multi-genre corpus, the Helsinki Corpus of English Texts (HC) (Rissanen et al. 1991) has also provided data for studies drawn on in this article. The HC, consisting of c.1.6 million words of English from c.700−1710, was compiled at the University of Helsinki under the supervision of Matti Rissanen in the 1980s (for further information, see Rissanen et al. 1993). Expanded versions have been provided by the University of Pennsylvania. Further materials used in historical sociolinguistics include A Corpus of English Dialogues (CED) (Kyto¨ and Culpeper 2006), covering the period 1560–1760, compiled at Uppsala University and Lancaster University under the supervision of Merja Kyto¨ and Jonathan Culpeper (for further information, see Kyto¨ and Walker 2003) and, for early American English, the Salem Witchcraft Records from 1692, re-edited by an international team (Rosenthal et al. 2009).
2 Early Modern England 1500–1700 It cannot be stressed too much that the pursuit of historical sociolinguistics requires a thorough knowledge of the conditions that prevailed in the society whose language is under examination. In order to provide some background information, this section gives a very brief description of Early Modern England, but a more detailed reconstruction of the society can found in Nurmi (Chapter 4), and in the relevant sociohistorical literature such as Laslett (1983), Wrightson (1991), Barry and Brooks (1994), Heal and Holmes (1994), Fletcher (1995), and Wrightson (2003). (For further references, see Nevalainen and Raumolin-Brunberg 2003).
45 Early Modern English: Sociolinguistics Ruled by the popular Tudor and Stuart monarchs, England underwent a change between 1500 and 1700 from the medieval three-estate society, involving clergy, nobility and laborers, into a country with a more complex social structure. Social historians have often described the divisions in hierarchical terms, as in Table 45.1. In line with what seems to be the usage among most social historians, this table and this article as a whole employs the early modern concept “rank” instead of the more controversial modern term “class”. Table 45.1: Rank and status in Early Modern England (after Laslett 1983: 38) Estate GENTRY
NON-GENTRY
Grade Nobility Gentry proper Professions
Royalty Duke, Archbishop, Marquess, Earl, Viscount, Baron, Bishop Baronet, Knight, Esquire, Gentleman Army Officer, Government Official, Lawyer, Medical Doctor, (wealthy) Merchant, Clergyman, Teacher, etc. Yeoman, Merchant, Husbandman, Craftsman, Tradesman, Artificer, Laborers, Cottager, Pauper
The nobility and the upper sections of the gentry proper formed the elite, comprising only a couple of per cent of the population. The upper clergy, the archbishops and bishops, found a place among the nobility because of their influential position. At the other end, laborers, cottagers and paupers occupied the lowest rungs of the social ladder. Although the main dividing line was usually drawn between the gentry and nongentry, this was not a rigid division. Beside a specific lifestyle, gentility involved land ownership and no need to work for a living, but the borderline could be crossed by people from the middle ranks. These consisted of people in administrative, legal and medical professions, wealthy merchants, and craftsmen. The social position of women was mainly derivative, unmarried women being categorized according to their fathers’ social position while the married ones followed their husbands. Education, too, was socially stratified. The way children were trained for adult life varied according to their social background and gender. Home was the most important place for teaching children the skills they needed in the future. The role of school attendance increased with time, but home tuition persisted. Children of the lower social strata rarely went to school and, if they did, their families could hardly afford it for more than a year or two. The curriculum of the educational institutions for boys, the grammar schools and beyond, was essentially classical, which means that it was predominantly boys from the upper and middling strata of society who acquired a command of Latin, the most prestigious language of the time. The educational system placed the language of law and administration beyond the reach of the lower ranks and women. The level of literacy was also socially stratified. Most of the estimates of literacy have been counted on the basis of the proportion of the population who signed their names instead of inscribing a mark on various public documents. This method may not have the best accuracy, but does apparently provide a confident estimate of those who were able to read, as reading was taught as a separate skill before writing. In the 17th century, most members of the gentry
717
718
V Early Modern English and professionals were able to read, whereas about only half of yeomen and craftsmen could do so. Women’s ability to read was at the same low level as that of male laborers. Despite the growth of migration within the country, dialectal variation in language was extensive. While it is possible to trace dialectal texts from the early part of the 16th century, it becomes more difficult to do so from later periods because of the increasing standardization of spelling, although this did not affect all private writing. On the whole, although considerable standardization of the written language took place during early modern times, no codified standard language existed before the 18th century.
3 The sociolinguistic patterns of language change 3.1 External constraints: gender, status, region and register One of the main aims of historical sociolinguistics is to find out how linguistic innovations spread among past populations. Researchers look for answers to questions such as who were the innovators and early adopters of a new form, was it men or women who first used this form, which part of the country did the innovation spread from, and so on. Although early modern society, described briefly above, changed over time, at a relatively abstract level we can consider it the independent variable against which linguistic variation is correlated (Chambers 2003: 18−19).
3.1.1 Gender Gender has proved to be one of the most robust external constraints in present-day sociolinguistics. In a seminal article, Labov (1990: 213−215) presents two basic principles of women’s participation in linguistic change: 1) women adopt prestige forms at a higher rate than men in linguistic change from above; 2) women use higher frequencies of innovative forms than men do in linguistic change from below. These two concepts, change from above and below, are used to account for social awareness in connection with the diffusion of innovative forms. Change from above refers to the importation of elements from outside the language system, e.g., from a foreign language, with full public awareness, while change from below refers to changes from within the language system without social awareness. The two general principles expressed by Labov, paradoxically involving opposing tendencies, suggest that women are more active in promoting linguistic change. As regards Early Modern England, it has become evident that Labov’s first principle of women adopting prestige forms before men does not hold because of women’s limited access to learning and classical education. On the other hand, many studies of Early Modern English show females leading changes in the same way as Labov indicates in his second principle, in change from below. Let us look at women’s participation in two early modern changes representing different areas of grammar. The personal-pronoun system underwent significant changes in Tudor and Stuart England, one of which was the replacement of the second-person singular pronoun thou/thee by ye/you. Furthermore, the old object form you rapidly replaced the subject form ye. This change took place circa 1480−1580, you spreading from the informal genres to the formal ones, thus representing a change from below (for details, see Raumolin-Brunberg 2005b).
45 Early Modern English: Sociolinguistics
719
Examples (1) and (2) illustrate the use of the second person subject pronoun in the letters of the merchant couple John and Sabine Johnson. While the husband relies on the traditional ye subject, his wife has adopted the new form you. (1)
trie owt this matter by examynyng of them eche alone by themselves that ye may knowe the trewthe, and then ye maie kepe and put from you whome ye thincke good, and that ye perseave to be fawlte (1546 John Johnson; CEEC)
(2)
Our Lord knowth, for I stand in doutt, wherefore I moest hartely desyre you to make all the sped hom that you can. And do you th[inc]ke, good husbond, what a great comfart it shal be to me to have [ \you\] here at that tyme (1546 Sabine Johnson; CEEC)
100%
80%
60%
40% Women Men
20%
0% 1520–1539
1540 –1559
1560 –1579
Figure 45.1: The gender distribution of the replacement of subject ye by you, 1500−1619 (CEEC, Nevalainen et al. 1998 and CEECSU, Kaislaniemi et al. 2000). Percentages of you
Figure 45.1, depicting the gender difference in the Corpus of Early English Correspondence from 1500−1619, shows that women were at least one generation ahead of men during the rapid change in the first half of the 16th century. Examples (3) and (4) illustrate another early modern change, the decline of multiple negation, also called double negation and negative concord. Sentential negation in early Middle English usually consisted of two parts, ne and not. In negative clauses, indefinites were expressed by their negative forms, e.g. nothing, and a sentential negator also occurred with no, never, etc. The first element of the sentential negator disappeared from most genres by 1600, and the negative indefinites were replaced by the assertive ones, e.g. forms beginning with any. Coordinate structures, typically including nor or neither, kept two negators for much longer (for details on multiple negation, see Nevalainen 2006). Example (3) illustrates the old usage with multiple negation, whereas (4) exemplifies the modern single negation with an assertive pronoun. (3)
and I shall not put you in no more troubul but I be sysch you hartly my Lorde that I may have it to morow at nyght at the farest (1516 Margaret Tudor; CEEC)
720
V Early Modern English (4)
I doe not hear of any man that dealeth so lustyly with the enemye as he doth (1586 Robert Dudley; CEEC)
Figure 45.2 shows the gender distribution of the percentages of single negation of all sentential negations under examination. It indicates that men led the decline of multiple negation throughout the 16th century. In men’s letters over half of the cases represent single negation as early as 1500–1539, while women still mostly resorted to multiple negation at this period. 100%
80%
60%
40% Men Women
20%
0% 1500–1539
1540 –1579
1580 –1619
Figure 45.2: The gender distribution of the decline of multiple negation, 1500−1619 (CEEC, Nevalainen et al. 1998 and CEECSU, Kaislaniemi et al. 2000). Percentages of single negation
The lack of a codified standard language in Early Modern England did not mean that prestige was not attached to some varieties, such as the language of the Royal Court and London with its surroundings (Puttenham 1589). However, this type of prestige was different from that of today’s standard and, on the whole, it seems that access to learned written language, very rarely open to women, made men leaders in early-modern changes from above, such as the above loss of multiple negation. It was not led by men in general but the leaders were well-educated upper-ranking and professional men who had access to learned types of writing (Nevalainen and Raumolin-Brunberg 2003). In fact, recent research has discovered more cases of female-led change than maleled change. In addition to the shift from ye to you, they include such well-known changes as the loss of the nasal from the possessives mine and thine, the introduction of the prop-word one, the replacement of the third-person singular suffix -th by -s (Nevalainen and Raumolin-Brunberg 2003: 118−125; for details see Section 3.1.2) and the rise of the auxiliary will at the expense of shall (Nurmi 2003b). As regards the male-led changes from above, apart from multiple negation we can mention inversion after initial negators (Nevalainen and Raumolin-Brunberg 2003: 129) and the rise of the meaning of ‘logical necessity’ of the auxiliary must (Nurmi 2003a). Recent studies have also identified changes that switch from male to female advantage. An interesting case is the use of periphrastic do in affirmative statements, a change in which men were ahead of women until c.1600, after which there was an increase in
45 Early Modern English: Sociolinguistics this regard led by women in the first decades of the 17th century. During the second half of the century, the difference between the sexes vanished and the use of do diminished in general. Nurmi (1999) suggests that the dropping of do in men’s language could be connected with changing prestige patterns, as the ascent of James I meant increasing Scottish influence on many matters of life, including language. Scottish English at the time used affirmative periphrastic do less than the southern varieties. In sum, there was considerable gender variation in the participation in ongoing changes in Early Modern England. Women led many changes from below as they do today, whereas changes from above, usually materializing as shifts following Latinate models, were led by men who had acquired a classical education.
3.1.2 Social stratification Social class is a constraint that is more or less taken for granted in the sociolinguistic research into present-day speech communities. There is every reason to believe that social stratification also played a role in the highly hierarchical early modern society. The creation of an appropriate stratification model for linguistic studies on the basis of the social hierarchies reconstructed by social historians (as in Table 45.1) is not without its problems. After several tests, a four-level model proved to be suitable for research into the material represented by the CEEC. The ranks are as follows: a. upper ranks: royalty, nobility, gentry and clergy; b. social aspirers: men of lower or middle ranks who entered the upper ranks by climbing at least two rungs on the social ladder; c. middle ranks: professionals and wealthy merchants; d. lower ranks: other non-gentry. The names of the categories speak for themselves, except that “social aspirers”, also called “upwardly mobile”, may need an explanation. Present-day studies have shown that upwardly mobile people are often very sensitive to linguistic attitudes so that they both tend to follow prestige patterns and to avoid stigmatized forms. Several trials on the CEEC have shown that the corpus is not large enough for a more fine-grained division. This is true for the lower ranks in particular, from which the data are inevitably limited because of their missing writing skills. The same is true of women, which is why the following case studies based on the CEEC only discuss the language of the male informants. Figure 45.3 shows the social rank distribution of the introduction of -s as the indicative third-person singular suffix during the early modern era, as its proportion grows from 8% to 93% in letters written by men. This change had been going on about five hundred years by the beginning of the early modern era, as the use of -s has been documented in the North of England as early as the 10th century. From there it spread to the South, where the corresponding suffix was -th. By the 15th century, -s had reached the late medieval London mercantile community, but -th was still the form the majority chose. As Figure 45.3 indicates, -s was favored by the lower ranks in the second half of the 16th century. A remarkable change took place around 1600. It is as if a stigma was lifted from -s, when its use increased from 19% to 58% between the twenty-year periods on each side of this dividing line. Even after 1600, the lower ranks had the highest score,
721
722
V Early Modern English 100%
80%
Upper Social aspirers Middle Lower
60%
40%
20%
0% 1500–1539 8%
1540 –1579 11%
1580 –1599 19%
1600 –1619 58%
1620 –1659 79%
1660 –1681 93%
Figure 45.3: Third-person singular suffix -s versus -th. Percentage of -s. Have and do excluded, male informants (CEEC, Nevalainen et al. 1998; after Nevalainen and Raumolin-Brunberg 2003: 144−145)
but the other ranks also chose -s in over half the cases. Examples (5)–(8) show the variation among the gentry in the 1620s. (5)
The French King geueth no sattisfaction to oe embassadors (1626 Nathaniel Bacon II; CEEC)
(6)
My sister giues you thankes for seending him to her (1625 Brilliana Harley; CEEC)
(7)
My brother’s monument goeth well forwarde (1628 Edward Bacon; CEEC)
(8)
the Counte Mansfeldes bysnis goes not forward as it aught (1625 Thomas Meautys I; CEEC)
It is interesting to see how quickly the usage of the social aspirers changed. It seems that these people sensed the changing attitudes towards the use of -s very early on. The above case study clearly shows that social stratification was an important factor in the diffusion of early modern linguistic changes. As far as other changes are concerned, the two shifts previously discussed in detail, the replacement of subject ye by you and the decline of multiple negation, also exhibit social stratification. Unlike the third-person -s, after its introduction in the middle ranks in Late Middle English, you spread very quickly into all ranks, including the upper ones. No stigma can have been attached to this change, and it is likely that the early adoption of you by the prestigious upper ranks guaranteed its rapid progress in general. Figure 45.4 shows the early modern distribution of the decline of multiple negation by four social ranks. This graph further illustrates the significance of male education referred to in the previous section. The leadership of this change falls on the social
45 Early Modern English: Sociolinguistics
723
100%
80%
Upper Social aspirers Middle Lower
60%
40%
20%
0% 1480–1519 22%
1520 –1559 61%
1560 –1599 88%
Figure 45.4: Use of single negation versus multiple negation 1480–1599. Percentage of single negation, male informants (CEEC, Nevalainen et al. 1998)
aspirers and middle ranks; in other words, people who were professionals in producing written documents, such as lawyers, administrators, and clergymen. Nevalainen and Raumolin-Brunberg (2003: 151) in fact show that even among the gentry those who held administrative offices or worked as lawyers used single negation more than other gentlemen. The lower ranks are clearly behind the others in the diffusion of this change, and this may well have been a reason for the later stigmatization of double negation. The Corpus of English Dialogues (CED) (Kyto¨ and Culpeper 2006), mostly used for socio-pragmatic research, also provides information about the role of the social rank in the diffusion of linguistic changes. This corpus, like in the Salem Witchcraft Records, contains material even from illiterate lower ranks, which is beyond the reach of the letter corpus. For example, in her study of the use of thou and you in Early Modern English dialogues, Walker (2007) shows that those in superior social position often addressed their social inferiors with thou but received you from them. This usage gradually diminished towards the 18th century, when you strengthened its position as the general pronoun of address. These descriptions of early modern changes testify that social stratification played a part in the diffusion of linguistic changes like that documented for present-day languages.
3.1.3 Region This section deals with the role regional variation played in the diffusion of changes, in other words, their supralocalization. As a matter of fact, in a VARBRUL analysis of several early modern changes, region proved more significant than gender and register in several cases (Nevalainen and Raumolin-Brunberg 2003: 199). Research has pinpointed two significant source areas of changes: the North and the London region. The importance of the North has a long history, as it is generally held that a large number of changes, such as the simplification of the inflectional system,
724
V Early Modern English originated in the North during the Old English period. The early modern developments can be considered a continuation of this protracted process. The significance of London, on the other hand, was a newer phenomenon, which was based on the enormous growth of this city during Tudor and Stuart times. Immigration to London created a community with considerable dialect mixing, the development of central administration made London and Westminster the place where the most important people lived, and the growth of the merchant community improved the city’s economy. All this raised London’s status in the eyes of ordinary Englishmen, and London English had all the prerequisites to become a prestige variety. We know, for instance, that in 1589 George Puttenham recommended the language of the “better-brought-up sort” of London and the Royal Court for aspiring poets. How, then, did the new forms spread in the country? The traditional hypothesis has been that diffusion takes place according to a wave model; in other words, changes gradually spread outwards from a centre. This is possible if populations are distributed relatively evenly and contacts are maintained in all directions. Geography and potential for communication can change the picture, however, so that mountains or rivers may hinder or promote contacts and linguistic changes with them. Another attested way of spreading linguistic innovations is dialect hopping. Trudgill (1986) showed that new forms can hop from one urban centre to another bypassing the countryside in between. It seems evident that in past societies it was migration that mostly accounted for this hopping phenomenon, and it is also clear that the capital was the centre that received a large number of immigrants from the countryside. However, people not only moved to London but many moved away from London after a stay of a few years and, of course, took the linguistic forms they had adopted with them. The regional diffusion of the innovative forms in the CEEC has been studied in four areas, namely London, East Anglia, the North (i.e. the counties north of Lincolnshire) and the Court. These areas should be self-explanatory except for the Court, which
100%
80%
London Court North East Anglia
60%
40%
20%
0% 1500–1539
1540 –1579
1580 –1619
1620 –1659
1660–1681
Figure 45.5: The replacement of -th by -s in verbs other than have and do, 1500–1681; regional distribution (CEEC, Nevalainen et al. 1998 and CEECSU, Kaislaniemi et al. 2000; excluding Sir Thomas Browne, whose language radically diverged from that of the other informants)
45 Early Modern English: Sociolinguistics
725
refers to the royal family and its courtiers, as well as diplomats and high administrative officers, many of whom lived in Westminster. Figure 45.5 describes a change that had its origin in the North. During the first decades of the 16th century, we only find -s in the letters written by northerners, although, as mentioned above, it had occurred in late-medieval merchants’ letters in London. The abrupt shift around 1600, already referred to in the previous section, is led by Londoners. During this period, East Anglia lags far behind. The following forty-year period testifies to a rapidly growing adoption of -s among courtiers and East Anglians alike. On the whole, here we see how a change with northern origin first spread to London and then from London to the countryside. The diffusion of subject you looks much simpler, as Figure 45.6 testifies. This is a typical change led by the London region, i.e., London and the Court. The curves for East Anglia and the North are almost identical. The spread of single negation as opposed to multiple negation was not very different from that of you, but the leading role of the Court was even more conspicuous. Its regional diffusion actually corroborates the results of the previous section, as courtiers represented the well-educated upper-rank gentlemen who were seen to lead this change. 100%
80%
London Court North East Anglia
60%
40%
20%
0% 1480–1519
1520 –1559
1560 –1599
Figure 45.6: The replacement of subject ye by you, 1480–1599; regional distribution. Percentages of you (CEEC, Nevalainen et al. 1998 and CEECSU, Kaislaniemi et al. 2000)
The significance of regional variation has also been attested in the dialogue corpus. Walker (2007: 102) shows that in witness depositions, thou was more frequent in the North – in other words, in areas where it may still occur in the dialect today.
3.1.4 Register and genre The situational variation of language use is reflected in register and genre variation, which both play an important part in the diffusion of linguistic changes. Studies of multi-genre corpora such as the Helsinki Corpus (Rissanen et al. 1991) have shown that changes from below are usually first attested in informal and speech-like genres. The Helsinki Corpus evidence indicates that subject you first appeared in oral genres (Raumolin-Brunberg 2005b: 64). As regards the introduction of the third-person suffix
726
V Early Modern English -s, Kyto¨ (1993: 124) shows that -s occurred in the private letters in the Early Modern section of the Helsinki Corpus before any other genre. On the other hand, testifying to a change from above, Rissanen (2000: 125) finds a relatively large number of occurrences of single negation in the Late Middle English legal texts in the Helsinki Corpus. This offers reason to assume that legal texts may have served as a model for the use of single negation, a practice that first spread among educated men. Letter data offers the possibility of using the interpersonal relationship as a basis for register variation and diffusion of new items. The division of letters into two groups, (1) letters exchanged between family members and friends and (2) letters between others, has shown that changes from below tend to be seen in the earlier group first. Raumolin-Brunberg and Nevalainen (1997: 711–713) observe that experienced writers, such as Sir Thomas More and Sir Thomas Wyatt, differentiated between family and non-family letters in the use of you and ye, among other changes. King Henry VIII’s letters form an excellent example of register variation between private and official writing. There are no occurrences of ye in his autograph love letters to Anne Boleyn, his second wife, although it occurs in his official letters written by secretaries. In the VARBRUL analysis of five changes, Nevalainen and Raumolin-Brunberg (2003: 199) show that register, analyzed on the basis of the relationship between the letter writers, tends to be less significant than gender and region. It is only in the earliest stages of the introduction of you that the group “family and friends” becomes the major factor promoting the use of you.
3.2 Idiolectal change A different view of the diffusion of linguistic changes is offered by examining the behavior of individual language users. There has been increasing interest in the way individuals participate in ongoing linguistic changes (see, e.g., Sankoff and Blondeau 2007). In a speech community, it is possible to find both people who participate and people who do not participate in changes in progress. The behavior of individuals under ongoing linguistic change is often discussed in terms of generational and communal change. Generational change refers to a situation in which there is idiolectal stability despite ongoing change in the community. In communal change, in turn, people change their language in adulthood, altering their language in the same direction. Labov (1994: 83–84) suggests that sound change and morphological change typically follow the pattern of generational change, while lexical and syntactic changes represent the converse pattern, communal change. Although this often seems to be the case, there is also evidence of lifespan changes, i.e. changes in adulthood, among sound changes (Sankoff and Blondeau 2007). It is worth pointing out that the well-known concept of apparent time goes back to the generational model, in other words, on the idea that adult speakers’ phonology and grammar are fixed. The apparent-time model is a convenient tool with which present-day sociolinguists can examine the diffusion of ongoing linguistic changes by comparing usage across generations of speakers in order to identify the direction and rate of change. Historical research has the advantage of following
45 Early Modern English: Sociolinguistics
727
changes in real time, in other words, how changes actually spread among populations and individuals. Figure 45.7 is a graphic presentation of the use of the third-person sibilant suffix in the language of nine individual CEEC informants over ten-year periods between 1570 and 1669. All of them have left letters spanning more than twenty years for posterity. At the same time as the figure testifies to extensive inter-individual variation, it also shows that six people changed the frequencies of the new and old forms over their lifespans, while three did not (for the numerical information on and background of the informants, see Raumolin-Brunberg 2005a). 100%
80% Hastings Francis Ralegh Walter Chamberlain John Harington John Gawdy Philip Holles John Howard Thomas Elizabeth of Bohemia Oxinden Henry
60%
40%
20%
0% 89
79
0–
7 15
0–
8 15
99
0–
9 15
19
09
0–
0 16
0–
1 16
29
0–
2 16
49
39
0–
3 16
0–
4 16
59
0–
5 16
69
0–
6 16
Figure 45.7: The percentage of third-person singular -s in nine idiolects, 1570–1669: a longitudinal study at 10-year intervals (CEEC, Nevalainen et al. 1998 and CEECSU, Kaislaniemi et al. 2000)
It is remarkable that there are two stable idiolects at the opposite ends of the scale: Sir Francis Hastings, who used the old form -th, and John Chamberlain, who chose the sibilant. Elizabeth, Queen of Bohemia, also exhibits high idiolectal stability during her long letter-writing career. The rest increase their use of -s at different rates, except that one informant, Philip Gawdy, both increases and decreases its use in a zig-zag pattern. The CEEC, with material from the same individuals over longer periods of time, makes it possible to study idiolects and their stability over time. At the same time as the findings so far show extensive inter-individual variation, they also testify to changes in idiolects in the lifespans of several language users.
3.3 Social networks The linguistic relationships within social networks form an interesting area of sociolinguistic research in their own right. Here we only raise issues that are connected with the diffusion of linguistic changes. According to James Milroy and Lesley Milroy (e.g. 1985), weak and uniplex ties in loose-knit social networks typically promote change, whereas the norm-enforcing character of networks with multiplex dense ties makes
728
V Early Modern English them resistant to change. On this view, linguistic innovators tend to be socially marginal people. On the other hand, on the basis of his Philadelphia data, Labov (2001: 385−411) observes that in the new and vigorous phase of change, the leaders of linguistic change are influential central members of their social networks. The analysis of social networks should be undertaken cautiously in historical linguistics because of the difficulties in the accurate reconstruction of networks. It seems that the network issue in Early Modern English can be approached in two ways: first, from the macro level, focusing on the general characteristics of the society in question and, second, by looking at particular individuals and their life-stories. The macro-level approach, based on the general history of Early Modern England, makes use of the fact that internal migration was common and even boosted by two major events, the Reformation in the 16th century and the Revolution with the Civil War in the 17th. As regards the Reformation, one might assume that only religious matters were at issue, but it had a broader effect on society. Among other things, the Church lost its riches and its former lands were sold to laypeople. Both the Reformation and the social turmoil of the Civil War increased weak ties in social networks and possibly accelerated the spread of ongoing linguistic changes (Raumolin-Brunberg 1998). Furthermore, London’s massive growth during the early modern era also led to an increase in weak-tie networks. Thus, it is only natural that the London region grew to be the centre from which linguistic changes spread to other parts of the country. The second approach, the examination of the social networks of particular individuals, is more difficult, because we seldom know enough of a person’s life to reconstruct their networks with sufficient accuracy. Some information has been gained by comparing the leaders of linguistic change at different phases of the diffusion process (Labov 1994:79–82; for the quantification, see Nevalainen and Raumolin-Brunberg 2003: 55). The CEEC data (Raumolin-Brunberg 2006: 130), for instance, suggest that during the incipient phase, when the new form was found at most in 15% of all the occurrences of the variable, the leaders of you and -s were geographically mobile people, probably with a great many weak links, which, according to Milroy and Milroy (1985), should promote the diffusion of linguistic change. On the other hand, the new and vigorous leaders, i.e. those who led the change when the new form, was found in 15–35% of the occurrences, seem to have been individuals with an influential social position, in other words, the kind of people Labov (2001: 385−411) speaks of as leaders of linguistic change. It is reasonable to suggest, for instance, that the rulers of the country like Henry VIII, leading the use of you, and Elizabeth I, an advanced user of third-person -s, were central figures in their social networks. The above suggests that the type of people who lead linguistic changes and their networks vary during the different phases of change. The analysis Milroy and Milroy (1985) offer seems to concentrate on the incipient phase, while Labov’s (2001: 385) arguments explicitly deal with new and vigorous changes. On the whole, despite being a less clear tool of analysis than the speaker variables such as gender, social status, age, etc., the examination of Early Modern social networks may help us understand the diffusion of innovative forms during this period.
45 Early Modern English: Sociolinguistics
4 Summary Despite the problems of accessing the lowest sections of society and women due to widespread illiteracy, studies of Early Modern English testify to the significance of gender, social rank, region, and register in the diffusion of linguistic change. We have seen that early modern women led a number of changes from below, whereas learned changes were led by men rather than women. Social stratification also proved significant, and it seems that, for rapid diffusion, a new variant needed to be accepted by the topmost ranks, as in the case of the third-person singular -s in the 17th century. The capital region, London and the Court, clearly became the centre from which several changes spread elsewhere in the country. Furthermore, changes from below first occurred in informal genres and registers. Some individuals changed their language across their lifespans, and weak-tie networks apparently promoted the diffusion of change. By using real-time evidence and linguistic material that represents genuine communication, sociolinguistic research on Early Modern English has enhanced our understanding of the ways grammatical changes diffuse among populations. Further research is needed to diversify and augment the picture that has been acquired so far.
5 References Barry, Jonathan and Christopher Brooks (eds.). 1994. The Middling Sort of People: Culture, Society and Politics in England 1550–1800. London: Macmillan. Biber, Douglas. 1988. Variation across Speech and Writing. Cambridge: Cambridge University Press. Chambers, J. K. 2003. Sociolinguistic Theory. 2nd edn. Oxford: Blackwell. Fletcher, Anthony. 1995. Gender, Sex and Subordination in England 1500–1800. New Haven/London: Yale University Press. Heal, Felicity and Clive Holmes. 1994. The Gentry in England and Wales, 1500–1700. London: Macmillan. Kaislaniemi, Samuli, Mikko Laitinen, Minna Nevala, Terttu Nevalainen, Arja Nurmi, Minna Palander-Collin, Helena Raumolin-Brunberg, and Anni Sairio. 2000. Corpus of Early English Correspondence Supplement (CEECSU). Department of English, University of Helsinki. http://www.helsinki.fi/varieng/CoRD/corpora/CEEC/ceecsu.html Kyto¨, Merja. 1993. Third-person present singular verb inflection in early British and American English. Language Variation and Change 5: 113−139. Kyto¨, Merja and Terry Walker. 2003. The linguistic study of Early Modern English speech-related texts: How “bad” can “bad” data be? Journal of English Linguistics 31(3): 221–248. Kyto¨, Merja and Jonathan Culpeper. 2006. A Corpus of English Dialogues 1560–1760. With the assistance of Terry Walker and Dawn Archer. Uppsala University and Lancaster University. http://www.engelska.uu.se/Research/English_Language/Research_Areas/Electronic_ Resource_Projects/A_Corpus_of_English_Dialogues/ Labov, William. 1990. The intersection of sex and social class in the course of linguistic change. Language Variation and Change 2: 205–254. Labov, William. 1994. Principles of Linguistic Change. Vol. 1: Internal Factors. Oxford/Cambridge, MA: Blackwell. Labov, William. 2001. Principles of Linguistic Change. Vol. 2: Social Factors. Oxford/Cambridge, MA: Blackwell. Laitinen, Mikko. 2007. Agreement Patterns in English: Diachronic Corpus Studies on CommonNumber Pronouns. Helsinki: Socie´te´ Ne´ophilologique. Laslett, Peter. 1983. The World We Have Lost – Further Explored. London: Routledge.
729
730
V Early Modern English Milroy, James and Lesley Milroy. 1985. Linguistic change, social network and speaker innovation. Journal of Linguistics 21: 339–384. Nevalainen, Terttu. 2006. Negative concord as an English “Vernacular Universal”: Social history and linguistic typology. Journal of English Linguistics 34(3): 257–278. Nevalainen, Terttu, Helena Raumolin-Brunberg, Jukka Kera¨nen, Minna Nevala, Arja Nurmi, and Minna Palander-Collin. 1998. Corpus of Early English Correspondence (CEEC). Department of English, University of Helsinki. http://www.helsinki.fi/varieng/CoRD/corpora/CEEC/index. html Nevalainen, Terttu and Helena Raumolin-Brunberg. 2003. Historical Sociolinguistics: Language Change in Tudor and Stuart England. London: Pearson. Nurmi, Arja. 1999. A Social History of Periphrastic do. Helsinki: Socie´te´ Ne´ophilologique. Nurmi, Arja. 2003a. The role of gender in the use of MUST in Early Modern English. In: Sylviane Granger and Stephanie Petch-Tyson (eds.), Extending the Scope of Corpus-Based Research: New Applications, New Challenges, 111–120. Amsterdam and New York: Rodopi. Nurmi, Arja. 2003b. Youe shall see I will conclude in it: Sociolinguistic variation of WILL/WOULD and SHALL/SHOULD in the sixteenth century. In: David Hart (ed.), English Modality in Context: Diachronic Perspectives, 89–107. Frankfurt: Peter Lang. Nurmi, Arja and Minna Palander-Collin. 2008. Letters as a text type: Interaction in writing. In: Marina Dossena and Ingrid Tieken-Boon van Ostade (eds.), Studies in Late Modern English Correspondence: Methodology and Data, 21–49. Bern: Peter Lang. Puttenham, George. 1589. The Arte of English Poesie. (English Linguistics 1500–1800, ed. by R. C. Alston, 110.) Menston: The Scolar Press, 1968. Raumolin-Brunberg, Helena. 1998. Social factors and pronominal change in the seventeenth century: The Civil War effect? In: Jacek Fisiak and Marcin Krygier (eds.), Advances in English Historical Linguistics, 361–388. Berlin/New York: Mouton de Gruyter. Raumolin-Brunberg, Helena. 2002. Stable variation and historical linguistics. In: Helena Raumolin-Brunberg, Minna Nevala, Arja Nurmi, and Matti Rissanen (eds.), Variation Past and Present: VARIENG Studies on English for Terttu Nevalainen, 101–116. Helsinki: Socie´te´ Ne´ophilologique. Raumolin-Brunberg, Helena. 2005a. Language change in adulthood: Historical letters as evidence. European Journal of English Studies 9(1): 37–51. Raumolin-Brunberg, Helena. 2005b. The diffusion of you: A case study in historical sociolinguistics. Language Variation and Change 17: 55–73. Raumolin-Brunberg, Helena. 2006. Leaders of linguistic change in Early Modern England. In: Roberta Facchinetti and Matti Rissanen (eds.), Corpus-based Studies of Diachronic English, 115–134. Frankfurt: Peter Lang. Raumolin-Brunberg, Helena and Terttu Nevalainen. 1997. Social embedding of linguistic changes in Tudor English. In: Raymond Hickey and Stanisław Puppel (eds.), Language History and Linguistic Modelling: A Festschrift for Jacek Fisiak on his 60th Birthday, 701–717. Berlin/ New York: Mouton de Gruyter. Raumolin-Brunberg, Helena and Terttu Nevalainen. 2007. Historical sociolinguistics: The Corpus of Early English Correspondence. In: Joan C. Beal, Karen P. Corrigan, and Hermann L. Moisl (eds.), Creating and Digitizing Language Corpora: Diachronic Databases. Vol. 2, 148–171. Basingstoke/New York: Palgrave Macmillan. Rissanen, Matti. 2000. Standardization and the language of early statutes. In: Laura Wright (ed.), The Development of Standard English, 1300–1800: Theories, Descriptions, Conflicts, 117–130. Cambridge: Cambridge University Press. Rissanen, Matti, Merja Kyto¨, Leena Kahlas-Tarkka, Matti Kilpio¨, Saara Nevanlinna, Irma Taavitsainen, Terttu Nevalainen, and Helena Raumolin-Brunberg. 1991. The Helsinki Corpus of English Texts. In: ICAME Collection of English Language Corpora (CD-ROM), 2nd edn., Knut Hofland, Anne Lindebjerg, and Jørn Thunestvedt (eds.), The HIT Centre, University of Bergen, Norway. For manual, see http://khnt.hit.uib.no/icame/manuals/hc/index.htm
46 Early Modern English: Pronouns
731
Rissanen, Matti, Merja Kyto¨, and Minna Palander-Collin (eds.). 1993. Early English in the Computer Age: Explorations through the Helsinki Corpus. Berlin/New York: Mouton de Gruyter. Romaine, Suzanne. 1982. Socio-historical Linguistics: Its Status and Methodology. Cambridge: Cambridge University Press. Rosenthal, Bernard, Gretchen Adams, Margo Burns, Peter Grund, Risto Hiltunen, Merja Kyto¨, Matti Peikola, Benjamin Ray, Matti Rissanen, and Richard Trask (eds.). 2009. Records of the Salem Witch Hunt. Cambridge: Cambridge University Press. Sankoff, Gillian and He´le`ne Blondeau. 2007. Language change across the lifespan: /r/ in Montreal French. Language 83(3): 560–588. Trudgill, Peter. 1986. Dialects in Contact. Oxford: Basil Blackwell. Walker, Terry. 2007. Thou and you in Early Modern English Dialogues: Trials, Depositions and Drama Comedy. Amsterdam/Philadelphia: John Benjamins. Wrightson, Keith. 1991. Estates, degrees, and sorts: Changing perceptions of society in Tudor and Stuart England. In: Penelope J. Corfield (ed.), Language, History and Class, 30–52. Oxford: Blackwell. Wrightson, Keith. 2003. English Society 1580–1680. London: Routledge.
Helena Raumolin-Brunberg, Helsinki (Finland)
46 Early Modern English: Pronouns 1 2 3 4
Early Modern English pronouns – an outline Personal pronouns Summary References
Abstract The chapter discusses the most important developments and changes of EModE personal and possessive pronouns and the explanations provided in the relevant specialist literature. As many of the pronoun changes were rather long-term developments, often originating in late Old English or Middle English, wherever necessary a short glance backwards or forwards into Modern English will be taken. The rise and the demise of the following variants is treated in detail: the third-person plural pronouns them and hem, the third-person singular neuter pronouns his and its, the first and second-person singular possessives my and thy vs. mine and thine, the second-person plural pronouns ye and you, and the second-person singular and plural pronouns thou and you. Methodologically, the EModE period (1500–1700) is broken down into three isolectal stages showing both the inventory and the use of forms at the beginning, in the middle, and at the end of the period.
1 Early Modern English pronouns – an outline Pronouns are grammatical (closed-class) words. On the one hand, this suggests that their number cannot be freely increased by processes of word-formation or borrowing. On the other hand, due to their discourse-deictic functions they are less likely than Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 731–743
732
V Early Modern English nouns to fall into disuse. Nonetheless, the EModE period saw the introduction of the possessive determiner its, the subject relative who and the loss of the thou (thee, thy, thine) paradigm. For reasons of space, the article deals with personal pronouns in detail, leaving aside all other types of pronoun. (For a concise outline of relative pronouns in Early Modern English, see Johansson, Chapter 49.) The introductions to Early Modern English by Barber (1997: 148–159), Go¨rlach (1994: 68–71) and Nevalainen (2006: 77–88) provide good overviews of the developments of the relative, interrogative, reflexive, intensive, demonstrative, and indefinite pronouns. Raumolin-Brunberg (1997) deals with reciprocal pronouns. The monograph by Heltveit (1953) deals with demonstrative pronouns, and van Gelderen’s (2000) book treats the history of English reflexive pronouns in detail.
2 Personal pronouns The most decisive change affected the second-person singular pronouns: the thou (thee, thy, thine) paradigm was replaced by the former plural forms you/yours/your. Apart from this major change, leading to the complete demise of the old singular forms in Present-day Standard English, a number of minor changes took place as well (see Table 46.1 below), involving in particular the following pronouns: – them and hem, ’em, with a northern form replacing a southern form, – his and its, with the new form its supplanting the older form his, – my vs. mine and thy vs. thine, resulting in a clear differentiation between determiner (my, thy) and independent function (mine, thine), – ye and you, leading to the generalization of the former objective form you.
Table 46.1: Early Modern English personal pronouns (from Nevalainen [2006: 77]; major changes indicated in boldface) Person/Number
Subjective case
Objective case
Possessive, determiner
Possessive, independent
1P SG 1P PL 2P SG
I we thou ~ ye → you ye → you he, she (h)it → it
me us thee ~ you
my/mine → my our thy/thine → thy ~ your your his, her his (thereof ) → its (of it) their
mine ours thine ~ yours
2P PL 3P SG personal 3P SG nonpersonal 3P PL
they
you him, her him, (h)it → it them (’em)
yours his, hers (his → its) theirs
The interesting question is: how was this reduction of variants brought about and which language-internal factors or external forces were at work? Hence, in the following sections these contrasts will be examined each in turn by describing the social embedding of the respective change and the language-internal factors that led to “streamlining” by sorting out the variants, as it appears from hindsight.
46 Early Modern English: Pronouns
2.1 The third-person plural pronouns them and hem The two forms them and hem for the third-person plural go back to different sources. The form hem stems from Anglo-Saxon, while the forms beginning with th- were borrowed from Old Norse in the OE period. The modern paradigm they/their/them is odd: an entire grammatical subsystem borrowed from another language. These come from Scandinavian þeir (nom) / þeirra (gen) / þeim (dat). This system was not, however, borrowed all at once; it took at least 400 years for the new paradigm to be established in the dialect complex that gave rise to the modern standards (Lass 2006: 74–75).
The substitution of the native forms by their Old Norse counterparts happened first in the north of England, where during the late OE period the contact with the Vikings in the area of the Danelaw had been most intense, presumably “because the Old English h-forms were ambiguous with singular pronouns” (Hope 2003: 91). During the ME period the Old Norse forms percolated from the north to the south, however, travelling at different speeds, with the nominative spreading faster than the possessive and the objective. “[W]e find the Northern dialect using the Scandinavian th-forms in all the three cases (nominative, genitive, and objective), the Midlands dialects using the thforms in the nominative but the h-forms in the genitive and objective, and the Southern dialect using the h-forms exclusively” (Brinton and Arnovick 2006: 275). Thus, in Late Middle English, for instance in Chaucer, we find nominative they, but hem and hir for the objective and possessive. “This process was completed in the fifteenth century, when the northern third-person plural object form them replaced the southern hem” (Nevalainen 2006: 78). Barber (1997: 151) is also of the opinion that the Old Norse forms “are the normal ones by 1500”, and Hope (2003: 91) remarks that the change was virtually complete in the south when Shakespeare was born in 1564. However, even for Shakespeare’s works Spevack (1968–80) lists 222 instances of ’em as opposed to 2,046 tokens of them. In the course of the period “hem becomes less frequent, and is rare in the seventeenth century, though it is still recorded as late as 1660. Its weak form was em, with the usual loss of initial /h-/ in unstressed syllables” (Barber 1997: 151). Nevalainen notes that “[t]he change would perhaps have been harder to detect in speech, because the unstressed forms of hem and them could have identical realisations, often rendered by ’em in writing imitating speech” (2006: 78).
2.2 The third-person singular neuter pronouns his and its In comparison to Modern English with the only forms it and its, Table 46.1 shows that for the third-person neuter pronouns a number of changes occurred in Early Modern English. In Old English, the neuter pronoun was declined hit, his, him, hit. During Middle English, unlike all other pronouns, the accusative and dative case forms were merged under the accusative rather than the dative form, turning the paradigm into hit, his, hit, so that at the beginning of the EModE period [t]he original form of the nominative and the accusative was hit, which was still in use in the sixteenth century. The word it is an example of the restressing of a weak form: it arose
733
734
V Early Modern English in ME when initial /h-/ was regularly lost in unstressed syllables. […] The disappearance of hit takes place during the sixteenth century, and by 1600 it is the normal form (Barber 1997: 150).
Sometimes these forms were even further reduced to ’t. However, “the ’t form has not survived in Modern English except in jocular use” (Brinton and Arnovick 2006: 332). By contrast, his remained the proper form of the possessive until the 1600s, making the forms for the possessive and determiner for the third-person masculine and neuter pronouns identical. This “ambiguous” use of his did not mirror the distinction between animate and inanimate referents. The clash between grammatical gender in contrast to notional or natural gender was felt by EModE speakers and was resolved in a number of ways. “Various substitutes were tried, clearly indicating a desire, conscious or unconscious, to avoid the use of his in the neuter” (Baugh and Cable 2002: 243), namely: – the use of the simple form it as a possessive (beginning in the 14th century), example (1): (1)
Wherfore I meruaile how our English tongue hath crackt it credit, that it may not borrow of the Latin as wel as other tongues […] (1586 Pettie, The Ciuile Conuersation of M. Stephen Guazzo; Go¨rlach 1994: 218) – periphrastic thereof and of it as alternative constructions, (2):
(2)
The Plantine trees also grow in that countrie, the tree is as big as a mans thigh, and as high as a firre pole, the leaues thereof be long & broade […] (1591 Hortrop, The Trauailes of an English Man; Go¨rlach 1994: 307) – the formation of its, by adding a possessive ending to it, (3):
(3)
And let confession make halfe amends, that euery language hath it’s Genius and inseparable form […] (1603 Florio, Montaigne’s Essays; Go¨rlach 1994: 219)
The new form spread rapidly. Barber (1997: 150–151) mentions that it was the normal form by the 1620s, and that the use of his as a neuter genitive became rare, but “lingered on until about 1670 […]. But once the form its had become firmly established, these other devices [see examples 1–3 above] became unnecessary”. However, “[a]t the beginning of the seventeenth century it was clearly felt as a neologism not yet admitted to good use. There is no instance of it in the Bible (1611) or in any of the plays of Shakespeare printed during his lifetime […]. Toward the close of the seventeenth century its acceptance seems to have gained momentum rapidly” (Baugh and Cable 2002: 244). Yet, this refers only to its function as a determiner. With reference to the grammar of Present-day English, Nevalainen (2006: 81) points out that “the independent possessive pronoun its is marginal even today”. In following Baugh and Cable (2002), we can thus conclude that If grammatical gender had survived in English the continued use of his when referring to neuter nouns would probably never have seemed strange. But when, with the substitution of natural gender, meaning came to be the determining factor in the gender of nouns, and
46 Early Modern English: Pronouns all lifeless objects were thought of as neuter, the situation was somewhat different (Baugh and Cable 2002: 243).
For a detailed treatment of the third person neuter possessive see Nevalainen and Raumolin-Brunberg (1994).
2.3 The first and second-person singular possessives my and thy vs. mine and thine By 1500, the pronoun-determiners have alternative forms: my or mine and thy or thine. By 1600, mine and thine have become less frequent, and by 1700 they are no longer used in this function. In short, the rules for the usage of the variant forms can be laid down as follows: At the beginning of the period in 1500, similarly to a/an, my and thy are used before consonants, and mine and thine before a vowel, or sometimes + vowel. In case of silent word-initial , as in honour, host or habit, we find mine/thine, and also before a pause. This usage changes in the 16th century. “By 1600, my and thy are almost without exception the forms used before consonants, while before vowels my and mine are in free variation, as are thy and thine: Shakespeare has both thine eyes and thy eye, both mine own and my own” (Barber 1997: 152). The following examples (4), (5) and (6) from Shakespeare’s plays (see Busse 2002: 224) underline this seemingly random patterning of variants within a single line of text, as they defy the neat distribution usually given in reference works: (4)
In thine own person answer thy abuse (1590/91 The Second Part of King Henry the Sixth II.i.40)
(5)
Throw thy glove, / or any token of thine honor else, […]. (1607/08 Timon of Athens V.iv.49–50)
(6)
I’ll take thy word for faith, not ask thine oath: […]. (1607/08 Pericles I.ii.120)
However, a systematic investigation of the Shakespeare corpus (see Busse 2002: Chapter 9) has revealed that by 1600 thy and thine were no longer in free variation before vowels, because “minimal pairs” show that meaningful choices were made on grounds of intra-textual constraints, in particular for stylistic reasons, as for euphony, emphasis, parallelism, antithesis, etc. According to Barber (1997: 152) mine and thine continue to recede during the 17th century, so that in standard literary prose my and thy can be regarded as the normal forms by 1700. Concerning the reasons for the ultimate demise of mine and thine, Go¨rlach (1994: 69) argues that in predicative use, the n-forms are characterized by greater intonational emphasis and their ability to occur before a pause or sentence boundary. Owing to this, the -n is generally retained. Thus, in the 17th century the Middle English phonetic distinction is understood grammatically. Go¨rlach also points out that the -n could be interpreted as a weakened form of one, as in such one, this one, which had become usual in this position after pronouns. Whether this really was the case, cannot be proved.
735
736
V Early Modern English Strang (1970: 139) argues along the same line: “since the use in final position was pronominal, the distribution could serve as matrix for a new, grammatical, distinction. The now familiar difference of use, +/n/ pronominal, -/n/ attributive, develops from this matrix at the end of the 16c; though in attributive use the old phonological distinction continued in use for a time”. Schendl (1997: 181) has brought forward the hypothesis that both “the synchronic and diachronic variation in the forms of the possessives might correlate with levels of formality and that the n-less variants entered the emerging written standard from the spoken language”. Schendl concludes that from the first half of the 16th century to the first half of the 17th century the stylistic value of the forms my/thy vs. mine/thine has resulted in a reversal of markedness. At the first stage, the distribution is basically phonologically determined, then my/thy are marked as informal, and finally these forms are unmarked, which leads to the marking of mine/thine as poetic. This view is also shared by Nevalainen, but she adds that this change – similar to the one of them replacing hem – has a regional dimension, too. “The loss of -n occurred earlier in the north than in the south. In the course of the sixteenth century the short forms my and thy spread to most contexts and the long ones were retained only in poetic language and fixed expressions (mine own, thine eyes)” (Nevalainen 2006: 78). Nevertheless, Wales (1996: 167–168) still includes thy and thine in the pronoun paradigm of present standard and non-standard English(es) “because they are still known and used by native standard English speakers as part of a general ‘elevated’ register”. For a concise recent case study of how the nasal in the first and second person possessives was lost, see Raumolin-Brunberg and Nevalainen (2007).
2.4 The second-person plural pronouns ye and you First examples for the use of the oblique form you functioning as a nominative can be found from the 14th century onwards, but at the beginning of the EModE period, there was still a clear distinction between the nominative and the objective forms of the second-person plural pronoun, with ye as the nominative and you as the objective form. By 1600, you takes over as the nominative form, making ye a less common variant. However, until the middle of the 16th century, ye still prevails in the nominative. In the second half of the 16th century the use of you as a nominative gains momentum, but the archaizing language of the Authorized Version of the Bible (1611) maintains ye in the nominative as in (7) and (8): (7)
No doubt but ye are the people, and wisdom shall die with you ( Job 12, 2)
(8)
The Lord deal kindly with you, as ye have dealt with the dead, and with me. The Lord grant you that ye may find rest […] (Ruth 1, 8–9; Pyles 1971: 202n.)
Kenyon (1914) has shown that in the Authorized Version of the Bible you only occurs as a nominative in 7% of all instances. Ye is also maintained in Tyndale’s translation of the New Testament. Simultaneously, many authors show vacillation in their use of the two variants. For instance, according to Wyld (1937: 330 quoted in Pyles 1971: 202n.), 16th-century authors like Lord Berners and Sir Thomas More clearly differentiated between both forms. Others like Roger Ascham, Bishop Latimer, Cavendish, or Lyly
46 Early Modern English: Pronouns (in Euphues) used both forms indiscriminately in the nominative, whereas Queen Elizabeth I used only you for both functions. Beginning in the third quarter of the 16th century, and coming to a close in the 18th, the paradigm is reordered with you as the majority form regardless of case. Ye is relegated to elevated literary usage such as poetic apostrophes. For example, in Shakespeare’s plays there occur a mere 343 tokens of ye in comparison to 22,222 tokens of you. Within these limits, ye frequently occurs in post-verbal position in imperatives and in sentences containing the optative subjunctive, in particular greeting formulae such as fare ye well. As concerns its discourse functions, ye is clearly marked as the affective pronoun in comparison to you, because it frequently occurs in exclamatives, which are mostly found as apostrophes in connection with abstract nouns together with either honorific or abusive vocatives (see Busse 2002: Chapter 10). A number of papers by Terttu Nevalainen and Helena Raumolin-Brunberg based on the Corpus of Early English Correspondence (CEEC) (Nevalainen et al. 1998) report on the social diffusion of the change. In these studies, the authors concentrate on extralinguistic variables such as social stratification, gender and apparent time. Apart from the social patterning of usage in the first subperiod (1520–1550) of the CEEC, the most striking result is that there are no instances whatsoever of ye in subject position in the second subperiod (1590–1620), “indicating that all the differences between the ranks observed some half a century earlier have been levelled out” (Nevalainen 1996: 66). In comparison to the other pronouns this development is rather exceptional, because ye and you are the only pronouns to have given up marking for nominative or oblique case, as this development did not take place for I-me, we-us, he-him, she-her, they-them (see Heltveit 1952: 378–379). Most often two explanations are given, a phonological and a syntactic, or rather a combination of the two, i.e. a confusion of weak forms in unstressed contexts and a cross-over analogy to the second person singular pronouns thou and thee. For example, Barber (1997: 149) says that “it is not entirely clear why it was the accusative form which became thus standardised, but it may have been by analogy with thou: the two forms usually had the same vowel in late Middle English, and […] the form [ jəʊ], rhyming with thou, still existed in the sixteenth century”. Strang also emphasizes this point: Phoneticians indicate that in the late 16c the strong form was /jəʊ/ (=PE* [Present-day English] /jaʊ/), weak /jʊ/. While the modern weak form is a continuation of the old one, the strong form is a new analogical creation, by lengthening of the weak one […]. The marked circumstances of the use of thou have, however, enabled the old strong form to survive, though the weak /ðʊ/ has been lost (Strang 1970: 140–141).
Go¨rlach (1994: 69) mentions the following factors: both ye and you had a common unstressed form [jə], which in addition to the redundancy of case marking led to wrong interpretations. This development received further strengthening through the cross-over analogy of vowels in thou and thee. He also points out that from the ME period onwards, the function of case marking has been largely supplanted by a fixed SVO-word order. With the consolidation of word order, the personal pronouns with their case marking are “over-characterized”, which made case shiftings without loss of understanding possible.
737
738
V Early Modern English According to Fries (1940: 201) the case differentiation of subject and object by means of word order rather than inflection makes progress during the 14th century, and is more or less fixed to SVO by 1500, although case marking in the pronouns is retained much longer. For the maintenance of this distinction in the pronominal system, Howe gives the following reasons: that morphologically the personal pronouns are […] on the whole portmanteau forms rather than suffixed inflection, and are thus phonologically less likely to lose inflection through the reduction of endings common in the Germanic languages in adjectives, nouns and verbs. Furthermore, the high degree of suppletion in the personal pronouns means that given phonological reduction, forms which have a suppletive distinction will tend to remain formally distinct longer than those with less suppletive distinctions, and this seems to be borne out by examples such as English ye-you […] (Howe 1996: 70).
A further reason that could have contributed to the confusion of forms is their spelling as both thou and you and thee and ye could be written with . Lass (1999: 154) offers the following short description and explanation for the replacement of ye by you, stressing that the process began in the fourteenth century especially in postverbal position, e.g. as subject of a preposed verb, as in (9): (9)
to morwe schal yow wedded be (1492 Guy of Warick; Mustanoja 1960: 125, my emphasis)
The spelling for you in unstressed positions and its pronunciation as /jə/ may have added to the confusion of the two forms. “Thus (at least in written language) there is an early precedent for confusing the shapes of the two forms. And the post-verbal use of you, even as subject, simply reflects the fact that oblique pronoun forms typically appear in this position as objects – a generalisation of linear position over syntactic function” (Lass 1999: 154). Lutz (1998: 197) has dealt with this case in detail. She argues that cross-over analogy, which may have had a partial effect on the switch from ye to you, cannot have been the decisive factor, because the singular nominative and oblique forms thou and thee do not show signs of confusion throughout the EModE period. For this reason, she instead assumes that the selection of the Middle English plural object form you as the only remaining form of direct address in standard English can be attributed to the interplay of various external and internal factors (Lutz 1998: 201).
2.5 The second-person singular and plural pronouns thou and you Whereas the previous changes have not received very much attention by scholars, the shift from thou to you has been studied in detail (see e.g. Finkenstaedt 1963). With the loss of singular thou (thee, thy, thine) in the standard form of the language and the generalization of you (your, yours), the number and case distinction got lost (see Table 46.1), thus creating “a notable asymmetry […] in the personal pronoun system” (Nevalainen 2006: 78). The gradual replacement of the singular forms began in the 13th century when the plural pronouns ye/yow spread as the polite forms in addressing one person. They “owe their new function most probably to French influence, particularly the courtly
46 Early Modern English: Pronouns literature” (Wales 1983: 108). From Chaucer’s time well into the EModE period, this contrast played an important role in marking socio-affective distance between interlocutors. The change from thou to you is most often explained in terms of sociolinguistic and pragmatic models or a combination of both. Brown and Gilman (1960) trace this use of pronominal address back to the Roman Empire of the 4th century, when the pronoun vos appears as a reflex of the nos pluralis majestatis of the emperor. From then on, they show the spread of the plural form and the development on two social planes. The vertical dimension of status yields the polite plural pronoun as a deferential address to superiors, and the singular pronoun as an address form for social inferiors. Discourse among social equals, when not intimate or well acquainted, favors a reciprocal exchange of plural pronouns, and the singular pronoun as a sign of intimacy. The main difficulties within the Brown and Gilman approach to the English language, and to Early Modern English in particular, lie in the fact that the overlay of social and affective usages of thou are often difficult or impossible to distinguish. Furthermore, in Shakespearean English there is often momentary fluctuation in pronoun usage among two interlocutors which cannot be explained by a model assuming static social hierarchies (see Wales 1983). For instance, Eagleson (1971: 13) treats such shifts in pronoun usage in terms of markedness: “you had become the unmarked or neutral form, while thou was the marked form, being used to register any important shift not simply in rank but especially in emotion, be it love or anger, respect or contempt”. He provides us with a striking example (10) taken from King Lear (discussed again by Stein 2003: 251). (10) Goneril [to Edmund]: Decline your head: this kiss, if it durst speak, / Would stretch thy spirits up into the air (1605 King Lear IV.ii.22–23) In trying to win Edmund over for her plans, Goneril tries to deceive him by making him believe that she is in love with him. She subtly achieves this effect by first addressing him with you, and after the bestowal of the kiss by passionately thouing him. The socio-pragmatic aspects of Shakespeare’s pronominal usage are further explored by Busse (2002), Stein (2003), and U. Busse and B. Busse (2010). Pronoun choices and in particular pronoun switches have been explained by microsociolinguistic or -pragmatic approaches. For example, Hope (1993: 85) argues that “rapidly modulating forms addressed to the same person” have hitherto been explained as being affectively marked deviations from the forms expected on a social basis according to the Brown and Gilman model of power and solidarity. Yet, his work on depositions would suggest that such shifts should, moreover, be interpreted on a micro-pragmatic level. Hope (1994: 58) finds the situation “with at least three competing systems – a social system, an emotional/politeness-based system, and a system in which ‘you’ is the only available form – all open to use by speakers” quite confusing. From these systems he only expects the last subsystem to show “socio-historical linguistic patterning consistently”, because the social and emotional subsystems “would be expected to disrupt it” (Hope 1994: 58).
739
740
V Early Modern English With Nevalainen (2006) we can summarize the general development of the change in the following way: Thou retreated to the private sphere, but could surface in public discourse when emotions ran high. Around 1600, thou is found in fiction, drama and poetry and in religious contexts of all kinds, especially with reference to God, as well as in trial records (Nevalainen 2006: 78).
However, the speed of the change, and even more importantly its interpretation, depend on the type of language studied. Taavitsainen (1997) discusses EModE genres and text types under the specific perspective of how personal affect is dealt with linguistically in fiction and adjoining text types. Among the factors she tests for personal affect are also the second person pronouns. From a diachronic statistical analysis, she finds that “the difference between the use of the second person singular in fiction and the adjoining genres is significant” (Taavitsainen 1997: 239), as fiction and to a lesser degree autobiography are the only genres to make use of thou throughout all phases of Early Modern English. However, in these text types the function of thou is different. “[T]he frequent use of thou in fiction is connected with the social classes depicted in these texts: country folk and lower and middle class people among which it was the unmarked pronoun of address […] In non-literary texts such as autobiography the use of thou to express heightened emotionality is evident” (Taavitsainen 1997: 256–257). For history, biography, diary and travelogue the figures for thou are almost negligible. On the basis of the CEEC, Nevala (2004) has studied the varying use of the pronouns of address in “letters written from one family member to another, as well as on correspondence between close friends” (Nevala 2004: 159), covering the time span from the early 15th century to the first half of the 18th century. In her own words, the main results of her study are summarized as follows: [N]o definite conclusions can be made about the influence of social rank on the use of thou. Instead, the writer’s social role and thus the power characteristics seem to affect the pronoun usage to a certain degree. […] Material from the 17th century shows an increase in the users of thou […]. The use of address pronouns also seems to be influenced to a certain degree by the situational level of intimacy and affection between members of a family and close friends. (Nevala 2004: 177, 178).
For further studies taking account of the social factors of the change from thou to you (especially in the 17th century), see Raumolin-Brunberg (1998; 2005). Walker (2007) has investigated the use of second-person pronouns in the three genres of Trials, Depositions, and Drama Comedy from 1560 to 1760 finding that [o]f the three genres, Trials had the lowest percentages of THOU, and there was a marked decline after 1600: THOU was rare in the seventeenth century, and did not occur after 1719 […]. The Depositions genre, by contrast, had much higher percentages of THOU. There was an unsteady decline across the 200-year period studied, but still as much as 21 per cent THOU in period 5 [1720–1760]. Drama Comedy exhibited a marked decline in THOU after 1640, and a further decline after 1720 (Walker 2007: 288–289).
46 Early Modern English: Pronouns As regards the factors influencing the use of thou and you, Walker (2007: 293) found “little evidence in the material that pronoun selection was determined by linguistic factors” such as different sentence and verb types or formulaic phrases. Instead, a wide range of extra-linguistic factors such as social distance, region, characterization, emotion, etc. “have been found to motivate pronoun usage in dialogues in the three speech-related genres” (Walker 2007: 292). In terms of sociolinguistic variables, “age did not seem a prime factor in pronoun selection” and “the influence of the sex parameter was less than clear” (Walker 2007: 292). These results show that the explanation for the change is not as simple and straightforward as it seems at first sight.
3 Summary The discussion of changes within the EModE system of personal pronouns shows that a number of minor changes and a fundamental change leading to the replacement of the singular pronoun thou (thee, thy, thine) took place. At the beginning of the EModE period in 1500, there is still an overlap in function between second-person singular and plural forms, but at the end of the period, about 1700, the only forms left over in Standard English are you and your. Viewed from a typological perspective, this change made the paradigm rather exceptional “not only in comparison with that of the immediately preceding stages of the language and with that of the other Germanic and Indo-European languages […], but also compared to many entirely unrelated languages” (Lutz 1998: 190), leaving Present-day Standard English with an asymmetrical pronoun system. All the pronoun changes are long-term developments, often beginning in Late Old English or Middle English. In the case of the replacement of hem by them, language contact with Old Norse was a decisive factor and the shift from thou to you has been attributed to the influence of courtly French literature in the ME period. While this shift seems to have been a socio-pragmatic matter, triggered by extra-linguistic factors, the change from ye to you in subject function shows an interplay of language-internal and external factors. The introduction of the non-personal third-person singular pronoun it resolved the conflict between the older system of grammatical gender and the newer system of notional gender. With my/mine and thy/thine a formerly phonological contrast is exploited for stylistic purposes before the nasal forms fall into disuse.
4 References Barber, Charles. 1997. Early Modern English. 2nd edn. Edinburgh: Edinburgh University Press. Baugh, Albert C. and Thomas Cable. 2002. A History of the English Language. 5th edn. London: Routledge. Brinton, Laurel J. and Leslie K. Arnovick. 2006. The English Language: A Linguistic History. Oxford: Oxford University Press. Brown, Roger W. and Albert Gilman. 1960. The pronouns of power and solidarity. In: Thomas A. Sebeok (ed.), Style in Language, 253–276. Cambridge, MA: MIT Press. Busse, Ulrich. 2002. Linguistic Variation in the Shakespeare Corpus: Morpho-syntactic Variability of Second Person Pronouns. Amsterdam/Philadelphia: John Benjamins.
741
742
V Early Modern English Busse, Ulrich, and Beatrix Busse. 2010. Shakespeare. In: Andreas H. Jucker and Irma Taavitsainen (eds.), Historical Pragmatics, 247–281. (Handbooks of Pragmatics, 8.) Berlin/New York: Mouton de Gruyter. Eagleson, Robert D. 1971. Propertied as all the tuned spheres: Aspects of Shakespeare’s language. The Teaching of English 20: 4–15. Finkenstaedt, Thomas. 1963. You and Thou: Studien zur Anrede im Englischen (mit einem Exkurs u¨ber die Anrede im Deutschen). Berlin: de Gruyter. Fisiak, Jacek, and Marcin Krygier (eds.). 1998. Advances in English Historical Linguistics. Berlin/ New York: Mouton de Gruyter. Fries, Charles Carpenter. 1940. On the development of the structural use of word-order in modern English. Language 16: 199–208. van Gelderen, Elly. 2000. A History of English Reflexive Pronouns: Person, Self, and Interpretability. Amsterdam/Philadelphia: John Benjamins. Go¨rlach, Manfred. 1994. Einfu¨hrung ins Fru¨hneuenglische. 2nd edn. Heidelberg: Winter. [English edition, Introduction to Early Modern English, Cambridge: Cambridge University Press, 1991]. Heltveit, Trygve. 1952. Notes on the development of the personal pronouns in English. Norsk Tidsskrift for Sprogvidenskap 16: 377–386. Heltveit, Trygve. 1953. Studies in English Demonstrative Pronouns: A Contribution to the History of English Morphology. Oslo: Akademisk Forlag. Hope, Jonathan. 1993. Second person singular pronouns in records of Early Modern spoken English. Neuphilologische Mitteilungen 94: 83–100. Hope, Jonathan. 1994. The Authorship of Shakespeare’s Plays: A Socio-linguistic Study. Cambridge: Cambridge University Press. Hope, Jonathan. 2003. Shakespeare’s Grammar. London: The Arden Shakespeare. Howe, Stephen. 1996. The Personal Pronouns in the Germanic Languages: A Study of Personal Pronoun Morphology and Change in the Germanic Languages from the First Records to the Present Day. Berlin/New York: Mouton de Gruyter. Kenyon, John S. 1914. Ye and you in the King James Version. PMLA Proceedings of the Modern Language Association of America 24, New Series 12: 453–471. Lass, Roger. 1999. Phonology and morphology. In: Lass (ed.), 56–186. Lass, Roger (ed.). 1999. The Cambridge History of the English Language. Vol. III. 1476–1776. Cambridge: Cambridge University Press. Lass, Roger. 2006. Phonology and morphology. In: Richard Hogg and David Denison (eds.), A History of the English Language, 43–108. Cambridge: Cambridge University Press. Lutz, Angelika. 1998. The interplay of external and internal factors in morphological restructuring: The case of you. In: Fisiak and Krygier (eds.), 189–210. Mustanoja, Tauno F. 1960. A Middle English Syntax. Vol. 1: Parts of Speech. Helsinki: Societe´ Ne´ophilologique. Nevala, Minna. 2004. Address in Early English Correspondence. Helsinki: Societe´ Ne´ophilologique. Nevalainen, Terttu. 1996. Social stratification. In: Terttu Nevalainen and Helena RaumolinBrunberg (eds.), Sociolinguistics and Language History: Studies based on the Corpus of Early English Correspondence, 57–76. Amsterdam: Rodopi. Nevalainen, Terttu. 2006. An Introduction to Early Modern English. Edinburgh: Edinburgh University Press. Nevalainen, Terttu and Helena Raumolin-Brunberg. 1994. Its strength and the beauty of it: The standardization of the third person neuter possessive in Early Modern English. In: Dieter Stein and Ingrid Tieken-Boon van Ostade (eds.), Towards a Standard English, 1600–1800, 171–216. Berlin/New York: Mouton de Gruyter. Nevalainen, Terttu, Helena Raumolin-Brunberg, Jukka Kera¨nen, Minna Nevala, Arja Nurmi, and Minna Palander-Collin. 1998. Corpus of Early English Correspondence (CEEC). Department of English, University of Helsinki. http://www.helsinki.fi/varieng/CoRD/corpora/CEEC/index. html
47 Early Modern English: Periphrastic
743
DO
Pyles, Thomas. 1971. The Origins and Development of the English Language. 2nd edn. New York: Harcourt Brace Jovanovich. Raumolin-Brunberg, Helena. 1997. Reciprocal pronouns: From discontinuity to unity. Studia Anglica Posnaniensia 31: 227–236. Raumolin-Brunberg, Helena. 1998. Social factors and pronominal change in the seventeenth century: The Civil War effect? In: Fisiak and Krygier (eds.), 361–388. Raumolin-Brunberg, Helena. 2005. The diffusion of you: A case study in historical sociolinguistics. Language Variation and Change 17: 55–73. Raumolin-Brunberg, Helena and Terttu Nevalainen. 2007. From mine to my and thine to thy: Loss of the nasal in the first and second person possessives. In: Ute Smit, Stefan Dollinger, Julia Hu¨ttner, Gunther Kaltenbo¨ck, and Ursula Lutzky (eds.), Tracing English through Time: Explorations in Language Variation, 303–314. Vienna: Braumu¨ller. Schendl, Herbert. 1997. Morphological variation and change in Early Modern English: my/mine, thy/thine. In: Raymond Hickey and Stanisław Puppel (eds.), Language History and Linguistic Modelling: A Festschrift for Jacek Fisiak on his 60th Birthday. Vol. 1: Language History, 179–191. Berlin/New York: Mouton de Gruyter. Spevack, Marvin. 1968–80. A Complete and Systematic Concordance to the Works of Shakespeare. 9 vols. Hildesheim: Olms. Stein, Dieter. 2003. Pronominal usage in Shakespeare: Between sociolinguistics and conversation analysis. In: Irma Taavitsainen and Andreas H. Jucker (eds.), Diachronic Perspectives on Address Term Systems, 251–307. Amsterdam/Philadelphia: John Benjamins. Strang, Barbara M. H. 1970. A History of English. London: Methuen. Taavitsainen 1997. Genre conventions: Personal affect in fiction and non-fiction in Early Modern English. In: Matti Rissanen, Merja Kyto¨, and Kirsi Heikkonen (eds.), English in Transition: Corpus-based Studies in Linguistic Variation and Genre Styles, 185–206. Berlin/New York: Mouton de Gruyter. Wales, Kathleen M. 1983. Thou and you in Early Modern English: Brown and Gilman reappraised. Studia Linguistica 37: 107–125. Wales, Katie. 1996. Personal Pronouns in Present-day English. Cambridge: Cambridge University Press. Walker, Terry. 2007. Thou and You in Early Modern English Dialogues. Amsterdam/Philadelphia: John Benjamins. Wyld, Henry Cecil. 1937. A History of Modern Colloquial English. 3rd edn. New York: Dutton.
Ulrich Busse, Halle/Saale (Germany)
47 Early Modern English: Periphrastic 1 2 3 4 5 6 7 8 9
DO
Introduction Origin Variation: discourse-based, social, and stylistic factors Variation: lexical and collocational effects Variation: phonotactic and syntactic factors Parametric change and the rise of DO Regulation of DO Envoi References
Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 743–756
744
V Early Modern English
Abstract In Early Modern English the modern use of auxiliary DO (Did you hear? I did not hear) largely replaced earlier full verb constructions (Heard you? I heard not). An affirmative without emphasis (I do hear = I hear) was also possible, as hardly today. This chapter discusses factors involved in the substantial variation characteristic of the period, possible motivations for the rise of DO, and the emergence of its modern profile. Variation corresponded to a wide range of parameters, including discourse and stylistic properties, the presence of particular lexemes and collocations, and the verb’s phonotactics. There were also major differences between clause types. These distributions may reveal motivations for the rise of DO, or simply reflect speakers’ choices during a more abstract change (the loss of V-to-I), proceeding at a shared “constant rate” across contexts. Many issues are unsettled, but electronic corpora provide a major opportunity for progress.
1 Introduction A distinctive property of today’s English is its use of DO in clauses negated by not and in inverted questions. In Chaucer’s period (the late 14th century) the finite lexical verb was used in such constructions, though DO was just beginning to appear. It gains over the finite lexical verb, and Early Modern English (1500–1700) is a period of substantial variation between the two, as finites of DO replace the earlier use of the finite lexical verb. So the playwright Jonson, writing c.1600, has both options available to him in the constructions of Table 47.1, and there is similar variation in negative questions. Table 47.1: Variation between with Affirmative inverted whquestions
Affirmative inverted yes–no questions
Negative declaratives with not
DO
and finite verb in Ben Jonson
DO
a. When did you see him? (1905–8 [1599] Jonson, Every Man out of his Humour 1643) b. O Madam, why do you prouoke your Father, thus? (1932 [1605] Jonson, Eastward Ho! 591, l. 161) a. Pray you, sir, did you see Master Fastidius Briske? (1905–8 [1599] Jonson, Every Man out of his Humour 2800) b. Doe you heare, sir? (1905–8 [1598] Jonson, Every Man in his Humour 2432) a. I am deafe, I doe not heare you; … (1932 [1605] Jonson, Eastward Ho! 610, l. 15) b. I look’t in the pot once, indeed, but I did not drinke. (1905–8 [1599] Jonson, Every Man out of his Humour 3933)
with finite lexical verb c. When saw you my neece? (1905–8 [1599] Jonson, Every Man out of his Humour 1284) d. And how approue you your sisters fashion? (1932 [1605] Jonson, Eastward Ho! 535, l. 132) c. O, Carlo! welcome: saw you Monsieur Briske? (1905–8 [1599] Jonson, Every Man out of his Humour 2748) d. How now Cutberd, succeedes it, or no? (1937 [1609] Jonson, Epicoene 196, l. 10) c. No sir, he saw him not. (1905–8 [1598] Jonson, Every Man in his Humour 271) d. I loue not your disputations, or your court-tumults. (1937 [1609] Jonson, Epicoene 258, l. 46) (Continued )
47 Early Modern English: Periphrastic
745
DO
Table 47.1: Continued with Negative imperatives with not
DO
a. Nay, doe not speake in passion so: … (1905–1908 [1598] Jonson, Every Man in his Humour 2707) b. doe not wrong the gentleman, and thy selfe too. (1905–8 [1598] Jonson, Every Man in his Humour 435)
with finite lexical verb c. speake not, though I question you. (1937 [1609] Jonson, Epicoene, 177, l. 7) d. Come, wrong not the qualitie of your desert, with looking downeward, couz; … (1905–8 [1598] Jonson, Every Man in his Humour 387)
At the beginning of Early Modern English DO was relatively uncommon, but by the end (1700) it was by far the more frequent variant in inverted questions, and had made substantial progress in negative declaratives and imperatives. In the major types of affirmative question its development followed an s-curve, being fastest in the middle of the 16th century as the curve crossed 50%, but negative questions and declaratives show a more complex development, with a major drop in incidence preceding a later recovery. In (and before) Early Modern English, DO also occurred rather freely in finite affirmative sentences like (1). This use of DO rose steadily to a peak of rather less than 10%, then declined. (1)
Affirmative declaratives with DO a. Sogliardo You will not serue mee, sir, will you? I’le giue you more then countenance. Shift Pardon me, sir, I doe scorne to serue any man. (1905–8 [1599] Jonson, Every Man out of his Humour 2217) b. George Your meat’s ready, sir, and your company were come. Carlo Buffone Is the loyne of porke enough? George I, sir, it is enough. Macilente Porke? heart, what dost thou with such a greasie dish? I thinke thou dost varnish thy face with the fat on’t, it lookes so like a glew-pot. (1905–8 [1599] Jonson, Every Man out of his Humour 3754)
Here there is no emphasis on the polarity as is normally required today. DO also occurs in sentences where there apparently is such an emphasis, (2), but these are in a minority. (2)
Afirmative declaratives with DO of emphatic polarity a. Sogliardo doe you know me, sir? Macilente I doe know you, sir. (1905–8 [1599] Jonson, Every Man out of his Humour 687) b. Clerimont No, sir, doe not take it so to heart: shee do’s not refuse you, but a little neglect you. Good faith, True-wit, you were too blame to put it into his head, that shee do’s refuse him. True-wit Shee do’s refuse him, sir, palpably, how euer you mince it. (1937 [1609] Jonson, Epicoene 191, l. 131)
746
V Early Modern English Thus at the beginning of the period, speakers and writers had a choice across a range of clause types: instead of using a finite lexical verb, they could use a finite of DO with the lexical verb in the infinitive. Various contextual, social, and stylistic factors apparently weighted this choice, but the two variants were clearly generally acceptable. By the end of the period, the predominant usage was approaching today’s, though change continued throughout the 18th century (Tieken-Boon van Ostade 1987). From 1700, most types of inverted questions strongly favored DO, and it was the more common (unmarked) choice with negative declaratives and imperatives. In affirmative declaratives it had become uncommon, but although examples which apparently show emphatic polarity increased in the 17th century, they seem to have remained a minority even towards the end of that century, so the modern situation in which affirmative DO ordinarily carries emphatic polarity or marks an implicit contrast is a later restriction. The situation can be illustrated from Ellega˚rd’s (1953) major investigation of Early Modern English prose works, by giving figures for three romances translated by Caxton from just before the period, five plays by Jonson in the middle, and Swift’s Journal to Stella just after the end (see Table 47.2). Table 47.2: Percentages of
DO
Affirmative inverted questions Negative declaratives Affirmative declaratives
in construction types in specific texts Caxton 1483–1489
Jonson, acted 1598–1609, printed 1600–1616
Swift 1710
20% (1/5) 3% (4/149) 2% (173/7200)
67% (231/344) 35% (47/133) 3% (138/4000)
93% (53/57) 87% (61/70) 0.2% (5/2800)
Here particular subcategories are omitted in accordance with Ellega˚rd’s (1953: 162) practice in his well known graph.
The major general problems raised by this construction are the following: – origin. How did it originate? What was the original process of grammaticalization? (Section 2) – distribution. What is the nature of the factors which structure the distribution of DO versus the finite full verb? Here there has been a wide spread of suggestions (phonotactic, syntactic, semantic, pragmatic, and discourse-based characteristics; the impact of processing; and social correlates). (Sections 3, 4, 5) – motivation. Why did DO make progress, and what is the nature of this change? Did any of the factors controlling its distribution also actually motivate the rise of DO? Or was it a more abstract type of change, in which speakers took choices because of the presence of variation without those choices impacting on the process of change itself? (Sections 5, 6) – regulation. A further major topic is the development of the modern distribution, in which DO has been characterized in terms of its use as a “last resort” expression of tense, occurring in constructions which require a finite auxiliary (inverted questions, clauses negated by not, clauses with emphatic polarity). In contrast, DO has been lost (or all but lost) in affirmative declaratives without emphasis, and substantial progress towards this position had been made by the end of the Early Modern period, as noted
47 Early Modern English: Periphrastic
747
DO
above. Why should the development of (Section 7)
DO
have been differentiated in this way?
2 Origin This will be only briefly discussed, since it belongs to the period before Early Modern English. The most persuasive account of the first stages of the grammaticalization of DO derives it from causative uses. These had a common variant which omitted the agent, and was focused on what happened, possibly being perfective. Thus in the Middle English example in (3) the question as to whether Grim made the baskets himself, or had them made is not relevant; the focus is on the fact that they were made (cf. the vagueness of modern “got” in the gloss). There are also contexts in which DO + infinitive is clearly causative, involving an agent who is not specified. (3)
Gode paniers dede he make, On til him, and oþer þrinne Til hise sones, to beren fish inne (c.1300 The Lay of Havelok the Dane 760–762; Skeat [ed., revised by Sisam] 1902) ‘He [Grim] got good baskets made, one for himself and another three for his sons, to carry fish in’
The construction became syntactically isolated, and was subsequently bleached and reanalyzed as containing an empty auxiliary (Denison 1993: 279–281, and see his summaries of the many competing theories of origin). Other suggestions include reinterpretation of DO + bare action noun, such as do helpe, do synne, do swynke – which faces the difficulty that the construction first appears in the south west where distinctive infinitival morphology of some weak verbs survived late – and conjectures that the construction has its source in second language acquisition in contact situations, in child language, or as a sporadic but recurrent form in Germanic dialect. There has also been a recent resurgence of interest in the relevance of contact with Celtic, which shows constructions combining DO with a verbal noun (see Filppula et al. 2008 for discussion and references). This may have involved a mutual development or a substrate influence on English, but it is difficult to draw confident conclusions in this area.
3 Variation: discourse-based, social, and stylistic factors 3.1 Discourse contrasts Since DO lengthens the verbal group, there is a potential stylistic contrast between sing and do sing etc., in which do sing is the marked choice. One clear effect is that affirmative DO is used for purposes of balance and sentence rhythm when a verb is final, as in (4), or when the verb would otherwise be inappropriately light, as probably in (5). (4)
in the high-street the Marchants and Tradesmen do dwell, … (1618 Taylor, Penniless Pilgrimage; Rissanen 1991: 331)
748
V Early Modern English (5)
The viij day of januarij dyd ryd in a care at westmynster the wyff of the grayhond … (1557 Machyn, Diary 8 Jan., entry 705; Bailey et al. [ed.] 2006).
In verse DO is clearly used to put an infinitive in rhyme. It has also been claimed that it is associated with a variety of effects involving focus or intensity (and these may underlie the later development of the modern use of affirmative DO for emphatic polarity). It is more common with verbs which may show strong emotional content, such as assure, believe, beseech, confess (Ellega˚rd 1953: 172), and Stein (1990), among others, claims that DO may mark intensity, and that it has discourse properties, marking prominence, such as that at the climax or peak of a narrative. It may also indicate a change of topic. DO is too general in Early Modern English for it to have had a precise semantic content, but it is the marked alternative in affirmative declaratives, and as such it has a range of marked uses.
3.2 Social and stylistic variation There is also clear differentiation across texts, some of which reflects social and stylistic variation. In the earlier part of the period, this bears on the question of whether the change is from above or from below. Ellega˚rd claimed that in the 15th century the use of affirmative DO was chiefly a feature of literary language, and this would be consistent with its being a top down change. But Nurmi’s results for the first half of the 16th century point to its being a change from below, and this seems more likely. She used data from the Corpus of Early English Correspondence (Nevalainen et al. 1998) and found that affirmative DO was more frequent in more informal family letters at this period and that social aspirers tended to avoid it (Nurmi 1999: 106, 189). This is also consistent with Rissanen’s (1991) convincing argument that the rise (and decline) of affirmative DO was initially a feature of the spoken language. In the second half of the century, however, it seems that DO had more prestige. It became more frequent in formal letters, and it seems to be particularly characteristic of the stylistically self-conscious Euphuistic writing of the mid-to-late 16th century. Gender is clearly a relevant (but somewhat puzzling) factor. Nurmi examined the period from 1580 in a sample from the Corpus of Early English Correspondence, noting that it shows a change from rise to decline in both affirmative and negative declaratives. She found a clear differentiation of gender between these clause types with a striking switch round. Women used significantly more DO than men in declaratives after 1620, but this reversed the situation found towards the end of the 16th century, when men used more DO than women (Nurmi 1999: 154, 171). There is also a striking drop in usage of DO not in negative declaratives seen in the second half of the 16th century in Ellega˚rd’s database, and from 1600 in the Corpus of Early English Correspondence. In both corpora there is a recovery later in the 17th century. Nurmi (1999: Chapter 10) suggests that this drop is due to the adoption of a new prestige model with a lower incidence of DO after James VI of Scotland became James I of England in 1603, since DO was diffusing northwards, and had a much lower incidence in northern English and Scottish English. Warner (2005), however, follows the dating of Ellega˚rd’s corpus, and proposes an earlier re-evaluation. He claims that Ellega˚rd’s database shows a smooth increase of use of DO not across time in texts with simpler lexis (many of which are plays), which he interprets as reflecting the vernacular
47 Early Modern English: Periphrastic
DO
development. But in texts with more complex lexis (including non-fictional prose, such as Milton’s Eikonoklastes), the level of DO not collapses dramatically towards the end of the 16th century, and shows little increase in the 17th. He suggests that this development may have had its roots in an avoidance in more carefully edited prose of the contracted form don’t, or the reduced forms which preceded it, given that there is evidence of some level of reduction from the 16th century.
4 Variation: lexical and collocational effects There were clear lexical effects. Some involve individual lexical items, and may be motivated by their properties. Dyd pryche ‘did preach’ is very common in Machyn’s Diary (1550–63) and it may well be that Machyn was avoiding the final consonant group [tʃt]; note that 40 years later Jonson included clutcht among the terrible vocabulary purged from Marston in the Poetaster (1601). Eat is also common with DO, especially in the preterit: in the Authorized Version (1611) did eat is the normal past, and ate is entirely absent in the Gospels (Ellega˚rd 1953: 179 note 1), presumably because of ambiguities caused by the fact that its ablaut had developed differently in different dialects and present and past were not always distinct. The best known lexical effect is that some particular verbs (Ellega˚rd’s 1953: 199 ‘know group’) are slow to occur with DO in negative declaratives: The group includes know, care, doubt, and fear, and we may add come and speak (201). There is a similar group of verbs in questions, including come, like, make, and say, but know is not especially uncommon with DO in questions. Such effects are often restricted to particular constructions, and collocations play an important role here. For example, I do well understand is common in the Cely Letters, and in Ellega˚rd’s database say is frequently found in How say you?, What say you?, while it is not so much the lexeme know itself that lags in its use of DO in negative declaratives in the 17th century, but the collocation I know not. Individual items can also be particularly frequent with DO (167). An example is think in negative declaratives in Ellega˚rd’s 17th century data, which is due mainly to the frequency of collocations with subject I: I do/did not think versus I think not, I thought not. It does not, however, seem to be the case that there is a pattern of lexical diffusion from a group of items which are in the lead (as has been claimed for some physiologically-motivated sound changes). Instead, DO seems to occur more frequently with less common verbs, but noticeably less frequently with a small group of particular lexical items, or particular collocations, as noted above. These are typically (but not invariably) verbs or collocations of high frequency. In this respect the spread of DO seems to be held back by the conservatism of such items, as has been noted for analogically-based morphological or phonological changes which affect the least frequent items first.
5 Variation: phonotactic and syntactic factors 5.1 Phonotactics and the motivation of change There is some evidence that DO tends to be used with subject thou when there would be a consonant group at the end of an inflected verb. This is particularly true of weak past tenses. Stein (1990) interpreted this as a major mechanism which motivated the rise of
749
750
V Early Modern English DO,
suggesting that the potential combination of a verb which had a consonant final stem with inflectional -st, -dst provided a trigger or motivation, and that DO was used “to circumvent undesired word-final consonant clusters” (171) on inflected main verbs. In his story, the raised use of DO was rapidly generalized to verbs with subject you, and subsequently to other categories. A difficulty with the major role assigned to this process as an engine of change is that the striking differences between syntactic categories remarked by Ellega˚rd are not adequately accounted for: Stein claims that questions are in the lead because they have “a greater sensitivity to phonotactic undesiredness” (Stein 1990: 156–157), but he simply takes this point for granted.
5.2 Syntax and the progress of change Syntactic factors which favor affirmative DO include the separation of subject and verb, notably by an adverb. But the most striking syntactic fact is the differences between major construction types, which can be substantial; see Table 47.2 above for some figures. Here are Ellega˚rd’s major categories, ordered in terms of their frequency of DO in the 15th and 16th centuries: negative inverted questions affirmative inverted questions (excluding questions with wh-object) negative declaratives affirmative declaratives negative imperatives Ellega˚rd (1953: 201–203) also noted that there was a strong preference for DO in transitive questions where the alternative with the finite full verb would have both subject and object after the verb, as in When saw you my neece or And how approue you your sisters fashion in Table 47.1 above. He referred to this as “one of the main inconveniences of inversion”, pointing out that questions with initial wh-object (What saw you?) or with intransitive verbs show a considerably lower level of DO. Kroch (1989a) proposed an account of this in terms of processing complexity. He suggested that by this date English speakers used a parsing strategy which initially interpreted V + NP as V + OBJECT. This would clearly have created a difficulty in inversion contexts where the object had not already been identified, and Kroch advanced an interpretation in which the high level of DO in such transitive questions reflected this parsing difficulty. This raised the possibility that change was driven by such considerations. He subsequently measured the rate of change for major construction types, supposing that competition between the incoming finite DO and the older finite full verb would result in an s-curve. He used an appropriate function (the logistic) which generated an s-curve, plotted the incidence of DO against time, and measured the rate of change for each context. He found that for the period up to 1575 the rates were not distinct (except for affirmative declaratives). This means that the s-curves for each context are the same shape. He called this identity the “Constant Rate Effect”. As a term this is potentially misleading, since clearly a change which proceeds along an s-curve starts slowly, speeds up, achieves a maximum rate of change, and then begins to slow down, and tails off. But the slope of any particular s-curve corresponds to
47 Early Modern English: Periphrastic
DO
a single coefficient within the equation type Kroch used, and when curves for different contexts have the same slope, this figure is constant. In order to account for this, he suggested that a single change (competition between finite full verb and DO) was taking place at an abstract level, and that surface distinctions between constructions were the result of various processing and discourse factors. Now note that, given the Constant Rate Effect, it is not clear that the construction types with the highest incidence of DO are driving the change. It had previously seemed natural to assume that the highest context would be the fastest (and the first), and this would clearly be the prime candidate for causation. But if all contexts proceed in parallel, none is “fastest”. Instead, there might be a single more abstract competition whose progress could be quite separately motivated; for example, as the increasing adoption of the unmarked value of the parameter for V-to-I movement (see below), or of a resulting structurally more economical grammar lacking this movement. Kroch comments that in such a case the pattern of favoring and disfavoring contexts does not reflect the forces pushing the change forward. Rather it reflects functional effects, discourse and processing, on the choices speakers make among the alternatives available to them in the language as they know it; and the strength of these effects remains constant as the change proceeds (Kroch 1989b: 238).
We might however suggest that such factors have a potential role in pushing the change forward, depending partly on our interpretation of the more abstract underlying change and its relationship to the rise of DO. This would fit with adaptationist views, in which language alters as strategies which are “fitter” to survive replace other less “fit” strategies. In biological evolution a single adaptation may have more than one advantage from the organism’s perspective, and we may imagine that the same will also be true where an abstract construction type is interpreted as the organism. Good candidates as contexts promoting change will be those which have the highest levels of DO, or which have the highest levels in a factor group which has a wide range of levels of occurrence, provided, of course, that there is a coherent set of interpretations of the value of the adaptation (the form with DO). From this perspective, both the parsing difficulty proposed by Kroch and the phonotactic problem noted by Stein may have contributed to the overall advance of DO across all contexts. Social factors (such as covert prestige) might also be implicated. There is good reason to think that the introduction of DO was a “communal” change in the sense that the community moved forward together, with individuals increasing their use of DO across time. An alternative would be for older individuals to retain an earlier level of usage, perhaps that set at acquisition, showing an “apparent time” effect across ages at any particular date. Ellega˚rd (1953: 166) notes for the 16th century “that there is a marked tendency for writers to use DO more frequently the older they grow”, and Stein (1990: 215–217) gives figures for Shakespeare from which it is clear that Shakespeare increased his usage of DO in questions with subject you across time. Warner (2004) examines Ellega˚rd’s data for questions, and finds that there is no support for an apparent time effect, but that the evidence is throughout consistent with communal change, under which individuals change their level of usage as they grow older. He goes on to suggest that this is consistent with a model of change based in usage, in which adults increase their level of use of DO in response to steady pressure from the kind of parsing difficulty discussed above.
751
752
V Early Modern English
6 Parametric change and the rise of
DO
Syntactic factors which have traditionally been thought to promote the rise of periphrastic DO include the general rise of the class of auxiliaries, and the associated higher frequency of periphrastic expression of tense, aspect, modality, and voice. Constructions with DO fit well into this general pattern. The rise of subject–verb–object unmarked order, as the earlier verb second order declined, may have promoted the rise of DO in inverted questions. Note that this sequence is retained when DO is used (DO–subject– verb–object), and it has indeed been suggested that an initial DO may have functioned as an interrogative marker. The position of DO before sentential negation has also been thought to provide a surface ordering (tense–negative–lexical verb) preferable to the alternative (tensed lexical verb–negative). Generativists have generally adopted a narrower focus, looking at the possibility that the rise of DO should be associated with changes in morphology and in the surface position of adverbs, which they account for more abstractly as the loss of V-to-I movement with full verbs. This process moves V from its position in VP to a higher functional position I, head of the inflectional phrase. This functional position encodes the tense and agreement contrasts which appear on the finite verb. In this higher position it precedes adverbs located at the edge of VP, so that they intervene between verb and object, as in (6). (6)
a. This most precious bloud that he shed on the Crosse, cryeth alwayes mercye for sinners … (a1535 Fisher, English Works 412; Mayor [ed.] 1876) b. These … confyrmed alwayes their lyues to the most holye lawes … of Chryste. (1544 Bale, A Brefe Chronycle concernynge Syr J. Oldcastell in Harl. Misc. [Malh.] I. 257; OED: s.v. conform)
But when V-to-I movement is lost, tense and agreement appear on the verb by the minimalist’s Agree or by a process of affix lowering, and adverbs located at the edge of VP no longer intervene between verb and object as in (7). (7)
a. Then I holde it best that we alwayes condempne The Byble readers. (1538 Bale, A Comedye concernynge Thre Lawes 1204; OED: s.v. bible). b. A most loving and carefull housholder, bicause he alwayes sent them rayne to prepare them foode. (1571 Golding, Calvin on Ps. lxviii. 10; OED: s.v. householder)
This change in verb–adverb ordering is clearly ongoing in late Middle English and Early Modern English (Ellega˚rd 1953: 182–186). If DO is taken to supply the position in I which is no longer occupied by a raised V, then the rise of DO can be seen as a reflex of the loss of V-to-I movement. The further question, as to what motivated this loss, has been answered by appeal to a parameter of Universal Grammar under which a language has to be sufficiently rich in verbal inflection for V-to-I to be acquired. Since Early Modern English shows a dramatic weakening of the earlier Middle English contrasts of person, number, and tense in the morphology of the verb, this might have led language learners to acquire a grammar which lacked V-to-I movement. There is a range of issues here, perhaps most pertinently the question as to why developments in Danish, Norwegian, and Swedish are so different. These languages are
47 Early Modern English: Periphrastic
DO
also held to have lost V-to-I movement but they permit NEG + verb outside verb second clauses, and retain verb + subject order in questions. Not + verb is found in Early Modern English, but only as a very uncommon option. The details of the loss of “rich inflection” are unclear, and there has been considerable discussion as to how the parameter should be interpreted. One suggestion has been that it was ultimately the loss of inflection for person in the past tense; another, the loss of co-occurring distinct morphemes for tense and person or number. The timing of these losses (and the related loss of infinitival morphology) also needs more detailed investigation. So too does the positioning of adverbs and the timing of the loss of verb–adverb order, which are not much investigated. Kroch makes the interesting claim that Ellega˚rd’s figures for the positioning of never show that the rate of loss matches that of the increase of DO, and that this implies the identity of the processes. But on Ellega˚rd’s figures, the loss of verb–adverb order is comfortably in advance of the rise of DO in the 15th and 16th centuries, and verb–adverb–object order seems to be lost by the late 16th century, though there is a substantial further period in which surface order in negatives could be interpreted as showing V-to-I movement. This might, however, be accounted for under hypotheses which split I into further categories, so that the loss of V-to-I movement is actually a sequence of losses of movement, as has been suggested.
7 Regulation of
DO
7.1 The decline of affirmative
DO
The term “regulation” refers to the development and establishment of the modern distribution of “last resort” DO, the latest stage in grammaticalization, in which DO becomes syntactically predictable. From 1650, DO is clearly moving towards this position. But it is perhaps possible to see the beginning of this development from the point when affirmative DO began to decline. There is an interesting difficulty over timing here, which reflects the nature of the texts being considered. In Ellega˚rd’s database affirmative DO seems to level off in the middle of the 16th century, and begin to decline towards the end of the century. But in the Helsinki Corpus the overall peak is later (Rissanen 1991), and in the Corpus of Early English Correspondence its decline is shown in “the first decades of the 17th century” (Nurmi 1999: 163). The Helsinki Corpus (Rissanen et al. 1991) is balanced between genres in a way not attempted by Ellega˚rd, and the Corpus of Early English Correspondence (Nevalainen et al. 1998) is based in a single genre. But it is interesting that Rissanen showed that the frequency of DO and timing of change in the Helsinki Corpus differed across genres, with trial transcripts and comedies (the text types closest to actual speech) showing a sharp decline from the 16th century, fiction remaining level, while other text types (including letters) continued to show an increase. This suggested to Rissanen that developments in speech and in writing were probably different, with speech losing affirmative DO earlier than writing, in which stylistic considerations supported its use in at least the first part of the 17th century (1991: esp. 328–329). Dating here then seems to depend on genre. Stylistic considerations were also involved in the decline of DO not (as discussed above): in Ellega˚rd’s database this is found in the second half of the 16th century, but in the Corpus of Early English Correspondence it can be seen at the beginning of the 17th century. Nurmi (1999: Chapter 10) suggested that the decline of DO in declaratives, both affirmative and
753
754
V Early Modern English negative, was a single sociolinguistic fact, due in both cases to the prestige of Scottish English, and she placed the grammatical differentiation of the two constructions well into the 17th century. But Rissanen’s findings that the genres closest to speech show an earlier decline of affirmative DO tend to suggest an earlier decline in spoken language. In negative declaratives there is indeed clear evidence (reviewed above) of a differential evaluation in the 17th century. But in the case of affirmative DO, one may wonder whether the decline follows not from a change in evaluation but from a grammatical difference.
7.2 Was there a grammatical change in the late 16th century? This has been posited by generative grammarians: it is potentially the point at which the modern system becomes discernible. It is also the point at which Kroch claimed that the loss of V-to-I was completed, since Ellega˚rd’s figures for the competition between the types of (6) and (7), where adverb position shows V-to-I or its absence, show that for never, V-to-I is found in only 3% of instances in 1575–1600 (Ellega˚rd 1953: 184). At this point, Kroch suggested, the different construction types became underlyingly distinct and no longer showed an identical competition between DO and the finite verb, so that the Constant Rate Effect ceased to hold. Kroch (1989b) believed that affirmative declaratives already had a lower coefficient of change than other contexts in the 15th and 16th centuries, and he interpreted this as the result of a competition between a different group of grammatical alternatives than were found in other contexts. A recalculation of the rate of change after some corrections to Ellega˚rd’s figures made by Warner (2006), however, shows that the rate is indeed actually the same in affirmative declaratives as in other contexts before the second half of the 16th century. This opens the possibility that a reanalysis at that point is not so much the going to completion of the loss of V-to-I, or of one of the components of the loss of V-to-I in a more articulated model of verbal projections, but instead the (general) adoption of the “last resort” status of DO, whereby DO is available only where other ways of realizing tense and agreement fail.
8 Envoi Despite the amount of work undertaken on DO, there is clearly much disagreement and major differences resulting from different theoretical positions. From the preceding discussion it is clear that many facts have been established, that much is yet to know, and that the history of DO is plentifully documented, so that it provides a wonderful resource for the investigation of long term linguistic change in a socially structured context. The properties of the variation between DO and finite verb clearly require substantial databases if they are to be studied profitably. Ellega˚rd (1953) collected a major database, and other prominent databases used in work on DO have been the Helsinki Corpus (Rissanen et al. 1991) and the Corpus of Early English Correspondence (Nevalainen et al. 1998), all mentioned above. The fact that these last two are publicly available in electronic form, the second in a parsed version, makes work on DO hugely easier, and we can expect future researchers to take advantage of these and other parsed corpora, including the Penn-Helsinki Parsed Corpus of Middle English (Kroch and Taylor 2000), the Penn-Helsinki Parsed Corpus of Early Modern English (Kroch et al. 2004) and the Penn Parsed Corpus of
47 Early Modern English: Periphrastic
DO
Modern British English (Kroch et al. 2010). Look forward to some decades of serious progress!
9 References Bailey, Richard W., Marilyn Miller, and Colette Moore. 2006. A London Provisioner’s Chronicle, 1550–1563, by Henry Machyn: Manuscript, Transcription, and Modernization. University of Michigan Press and the Scholarly Publishing Office of the University of Michigan University Library. http://quod.lib.umich.edu/m/machyn/ Denison, David. 1993. English Historical Syntax: Verbal Constructions. London/New York: Longman. Ellega˚rd, Alvar. 1953. The Auxiliary ‘Do’: The Establishment and Regulation of its Use in English. Stockholm: Almqvist & Wiksell. Filppula, Markku, Juhani Klemola, and Heli Paulasto. 2008. English and Celtic in Contact. New York/London: Routledge. Jonson, Ben. 1905–8 [1598]. Every Man in his Humour. In: Willy Bang (ed.), Ben Jonsons Dramen. Materialien zur Kunde des a¨lteren Englischen Dramas. Vol 7. Louvain: A. Uystpruyst. Jonson, Ben. 1905–8 [1599]. Every Man out of his Humour. In: Willy Bang (ed.), Ben Jonsons Dramen. Materialien zur Kunde des a¨lteren Englischen Dramas. Vol 7. Louvain: A. Uystpruyst. Jonson, Ben. 1932 [1605]. Eastward Ho! In: C. H. Hereford and Percy Simpson (eds.), Ben Jonson. Vol. 4. Oxford: Clarendon Press. Jonson, Ben. 1937 [1609]. Epicoene. In: C. H. Hereford and Percy Simpson (eds.), Ben Jonson. Vol. 5. Oxford: Clarendon Press. Kroch, Anthony. 1989a. Function and grammar in the history of English: Periphrastic do. In: Ralph W. Fasold and Deborah Schriffin (eds.), Language Change and Variation, 132–172. Amsterdam/Philadelphia: John Benjamins. Kroch, Anthony. 1989b. Reflexes of grammar in patterns of language change. Language Variation and Change 1: 199–244. Kroch, Anthony and Ann Taylor. 2000. Penn-Helsinki Parsed Corpus of Middle English. 2nd edn. http:www.ling.upenn.edu/hist-corpora/PPCME2-RELEASE-3/index.html Kroch, Anthony, Beatrice Santorini, and Ariel Dietani. 2004. Penn-Helsinki Parsed Corpus of Early Modern English. http:www.ling.upenn.edu/hist-corpora/PPCEME-RELEASE-2/index. html Kroch, Anthony, Beatrice Santorini, and Ariel Dietani. 2010. Penn Parsed Corpus of Modern British English. http:www.ling.upenn.edu/hist-corpora/PPCMBE-RELEASE-1/index.html Mayor, John E. B. 1876. The English Works of John Fischer, Bishop of Rochester. (Early English Text Society, E.S., 27.) London: N. Tru¨bner. Nevalainen, Terttu, Helena Raumolin-Brunberg, Jukka Kera¨nen, Minna Nevala, Arja Nurmi, and Minna Palander-Collin. 1998. Corpus of Early English Correspondence (CEEC). Department of English, University of Helsinki. http://www.helsinki.fi/varieng/CoRD/corpora/CEEC/index. html Nurmi, Arja. 1999. A Social History of Periphrastic DO. Helsinki: Socie´te´ Ne´ophilologique. Rissanen, Matti. 1991. Spoken language and the history of do-periphrasis. In: Dieter Kastovsky (ed.), Historical English Syntax, 321–342. Berlin/New York: Mouton de Gruyter. Rissanen, Matti, Merja Kyto¨, Leena Kahlas-Tarkka, Matti Kilpio¨, Saara Nevanlinna, Irma Taavitsainen, Terttu Nevalainen, and Helena Raumolin-Brunberg. 1991. The Helsinki Corpus of English Texts. In: ICAME Collection of English Language Corpora (CD-ROM), 2nd edn., eds. Knut Hofland, Anne Lindebjerg, and Jørn Thunestvedt, The HIT Centre, University of Bergen, Norway. For manual, see http://khnt.hit.uib.no/icame/manuals/hc/index.htm Skeat, Walter William. 1902. The Lay of Havelok the Dane. 2nd edn revised and corrected by Kenneth Sisam, 1915, with further corrections 1956. Oxford: Clarendon Press.
755
756
V Early Modern English Stein, Dieter. 1990. The Semantics of Syntactic Change: Aspects of the Evolution of ‘do’ in English. Berlin/New York: Mouton de Gruyter. Tieken-Boon van Ostade, Ingrid. 1987. The Auxiliary Do in Eighteenth-century English: A Sociohistorical-linguistic Approach. Foris: Dordrecht. Warner, Anthony. 2004. What drove DO? In: Christian J. Kay, Simon Horobin, and Jeremy Smith (eds.), New Perspectives on English Historical Linguistics, Vol 1. Syntax and Morphology, 229–242. Amsterdam/Philadelphia: John Benjamins. Warner, Anthony. 2005. Why do dove: Evidence for register variation in Early Modern English negatives. Language Variation and Change 17: 257–280. Warner, Anthony. 2006. Variation and the interpretation of change in periphrastic do. In: Ans van Kemenade and Bettelou Los (eds.), The Handbook of the History of English, 45–67. Malden: Blackwell.
Anthony Warner, York (UK)
48 Early Modern English: The Great Vowel Shift 1 2 3 4 5
Introduction Why “Great Vowel Shift”? On the history of Great Vowel Shift theories Motivating the Great Vowel Shift and avenues for further research References
Abstract The long-vowel inventories of all modern English accents and dialects differ substantially from the pronunciations that existed around 1400. Most of the relevant changes have been described as being interlinked and part of the so-called “Great Vowel Shift” (GVS), but consensus in the pertinent scholarship is limited. This chapter pursues a discussion of the history of GVS theories, the major issues and arguments. It is seen that some longstanding tenets and theories have a weak foundation and that the GVS is well known only in the sense that it is widely known. Despite a vast literature, many aspects of the changes are still poorly understood and, probably because of the vast literature, most aspects are controversial. Some of the controversies, however, turn out to be definitional rather than factual in nature. In this context, this chapter provides the likely paths of development from Middle to Modern English(es).
1 Introduction A handbook article on what has occasionally been called the “watershed” of the history of English phonology must aim at broad coverage and focus on what is common ground. The problem for a chapter on what is traditionally labelled the “Great Vowel Shift” is that few things have remained undisputed in the literature, for this is probably the most-written-about development in the history of the English language. A focus on common ground is thus virtually impossible, as is an exhaustive treatment. Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 756–776
48 Early Modern English: The Great Vowel Shift And although some of the recent literature seems to converge on the position that only the changes which affected the Middle English (hereafter ME) phonemes /iː/, /uː/, /eː/, and /oː/ are interrelated, and thus part of a shift (a shift which would then not be so great after all and perhaps not merit capitalization and a definite article), this position may turn out to be ephemeral. For the sake of completeness and to give due attention to older accounts, this chapter will discuss all long vowels and thus include also the developments of the lower half of the vowel space, i.e. the developments that affected ME /ɛː/, /ɔː/, and /aː/. The discussion will start with the uncontroversial, proceed to majority views, and conclude with a treatment of conflicting theories. The label “Great Vowel Shift” was introduced by Otto Jespersen (1909) almost exactly 100 years ago. But rather than offer yet another review of the literature (see McMahon 2006a or 2006b for a recent synopsis) and rather than present new or more detailed phonetic facts or conjectures based on individual writers’ orthoepistic and textual evidence, this chapter will adopt a more global perspective. It will try to shed some new light on the discussion by (a) re-evaluating Dobson’s (1968) interpretations of a range of 16th and 17th century sources in a quantitative manner, and by (b) looking at the contemporary intellectual background and the assumptions behind certain theories. In particular, the chapter will focus on the early theories by Jespersen and Luick (1896) from the turn of the 20th century, but also comment in passing on later biology-driven analogies and structuralist theories. Perhaps not surprisingly, it will turn out that the answers to the most fundamental questions – whether the label “shift” and the epithet “great” are appropriate – hinge crucially on a researcher’s perspective or definition and that different perspectives entail different merits and problems. In the course of this chapter, some of the classic problems, listed as (i) to (v) below, will therefore lose their poignancy, but new problems will arise. The article will conclude with a discussion of motivations and potential avenues for future research.
2 Why “Great Vowel Shift”? In the past three decades, research on the series of changes known as the “Great Vowel Shift” has centered on counterexamples and focused on why what happened to the ME long vowels should not be considered “great” or a “shift”. This chapter will begin with a defense of the traditional label, although it is by no means the first to do so. In another recent handbook article, McMahon (2006a) discusses in a systematic way the classic and partly interrelated five “problems” identified by Lass (1976) and Stockwell and Minkova (1988), around which most of the literature revolves: (i) Inception: where in the vowel space did the series of changes begin? (ii) Order: what is the chronology of individual and overlapping changes? (iii) Structural coherence: are we dealing with interdependent changes forming a unitary overarching change or with local and independent changes? (iv) Mergers: is the assumption of non-merger, i.e. preservation of phonemic contrasts, viable for language change in general and met in the specific changes of the GVS?
757
758
V Early Modern English (v) Dialects: how do we deal with dialects which did not undergo the same changes as southern English or in which the changes proceeded in a different order? After careful consideration of the issues and evaluation of the previous literature, McMahon concludes that while there is no simple answer to any of the above problems, the label “Great Vowel Shift” is justified beyond aesthetic and didactic grounds, certainly for the upper half, but probably also for the lower half, of the vowel space. The analyses offered in the present account will essentially confirm this position.
2.1 Why “great”? In the late 19th century, linguists like Luick (1896: 306–307) were struck by the fact that all long vowels of the English spoken around Chaucer’s time changed qualitatively in subsequent centuries. And the qualitative changes were so significant that for 17th century pronunciations new phonemic labels are necessary in order to avoid crude misrepresentations of the phonetic facts, certainly (but not only) for the predecessors of modern southern British English. For convenience and familiarity among the expected readership, my first reference point will be the accent that is referred to as “Received Pronunciation” or “RP” in its Present-day English (PDE) form, which – although supposedly supraregional – is essentially based on the pronunciation of educated southern British English speakers (see Volume 2, Mugglestone, Chapter 121). Table 48.1 lists all ME long vowels and their PDE RP reflexes. Lexical exceptions as well as dialects and accents other than RP will be dealt with in later sections. Table 48.1: Modern RP pronunciations of the ME long vowels with PDE orthographies (“C” stands for “consonant”; adapted from Barber 1997: 105) Middle English
Modern English (RP)
example
typical (and rarer) PDE spellings examples iCe, -y, -ie, (i+ld; i+nd) tide, fly, pie (child, kind) ou, ow mouse, how ee, ie seed, field oo, (oCe, -o) food, (move, who) ea, ei, eCe heath, conceit, complete oCe, oa, (-o, oe) hope, boat, (so, foe) aCe make, dame
(I)
iː
>
aɪ
time
(II)
uː
>
aʊ
house
(III)
eː
>
iː
see
(IV)
oː
>
uː
boot
(V)
ɛː
>
iː
sea
(VI)
ɔː
>
əʊ
sole
(VII)
aː
>
eɪ
name
It is true that there exist northern English and Scottish dialects that have not participated in all of the changes sketched above. And yet the vast majority of modern native
48 Early Modern English: The Great Vowel Shift speakers of English worldwide have pronunciations that diverge in relatively minor ways from modern RP, notably so when their varieties are compared to early Middle English (that is, pre-GVS) pronunciations. In fact, many modern dialects can be shown to be conservative relative to RP and can thus be located somewhere on the paths from ME to RP (whose intermediate stages are specified in Table 48.2 and Figure 48.1 below). Consider, for instance, Edinburgh English dialects which are currently diphthongizing their reflexes of ME /uː/, /ɛː/ and /ɔː/ (Schu¨tzler 2009). This, of course, does not mean that RP is more advanced in the sense of “being superior” or even a natural endpoint of diatopic or diastratic variation, as is immediately obvious from the fact that modern RP speakers – similar to Australian and New Zealand English speakers – are diphthongizing /iː/ again in words like see, me, tea. Just how complex the situation is can be seen in American English, which varies between [oː], [o] and [oʊ] for ME /ɔː/ in words like go and goat: depending on the history of a dialect, the monophthongal variants [oː] and [o] can be either progressive (i.e. monophthongizations of [oʊ]) or conservative (i.e. reflect one-step raisings from ME /ɔː/, as in most modern Scottish and Irish English dialects outside Edinburgh and Dublin; see also Section 3 below for discussion). In any case, such evidence lends further support to the uniformitarian hypothesis (see Christy 1983), which most modern research on phonetic and phonological change is based on and according to which changes that are impossible today were impossible in the past because the same principles hold for changes irrespective of the period during which they occur. Lass (1997: 24–32) offers an illuminating updated account of the uniformitarian hypothesis, including the Uniform Probabilities Principle, which states that “the (global, cross-linguistic) likelihood of any linguistic state of affairs (structure, inventory, process, etc.) has always been roughly the same as it is now” (Lass 1997: 29). From this follows that present-day changes are in principle no different from historical ones and may thus shed light on the past. This chapter therefore considers conservative as well as progressive dialects if they exhibit changes that may enhance our understanding of the GVS. Let us leave aside for a moment the question of whether or not the changes in Table 48.1 are interlinked and thus merit the label “shift” (for discussion, see Sections 2.2 and 2.4 below). Allowing for some simplification – as all models, theories, and handbook articles must – the changes involved certainly meet the criteria for a number of strong labels in historical phonology. In the dialects that participated in the shift almost the entire English lexicon was affected by the changes in (I) to (VII). In other words, whatever the individual histories and intermediate stages, it is obvious that it was essentially phonemes that changed. We can thus label each individual change without oversimplifying too much an “unconditioned”, i.e. “context-free” sound change that deserves to be called a “neogrammarian” sound change – though not in the strongest form of the hypothesis, which claims that sound change affects all words and all speakers of a speech community simultaneously, because some items (like do, good) were affected by the changes earlier than others (cf. Ogura 1987; Lass 1999: 78 and the discussion in Labov 1994: Chapter 17 on sound change vs. lexical diffusion). Indeed, precisely the fact that some exceptions to the GVS can be explained by the existence of phonetic variants underpins the neogrammarian label: low-stress items like and or my [mɪ], as in me mum, for instance, simply had no long vowel because high frequency
759
760
V Early Modern English and low stress lead to vowel lenition (cf. Bybee 2003); and differences like sane vs. sanity or divine vs. divinity display a regular pattern, too (cf. McMahon 2007). It is at this early point that a chapter on the Great Vowel Shift must leave the comfortable ground of unanimous scholarly consensus and enter the field of majority views because the phonetic details or developmental paths with intermediate stages that have led to modern English RP are not uncontroversial, although even here the differences in opinion are smaller than they seem at first sight. Different symbols like [iy, ɪi, ɪj] today often do not represent differences in views on the phonetic facts but are explainable in terms of different transcription traditions and conventions. Bloch and Trager (1942) as well as Trager and Smith (1951) systems from the 1940s and 1950s (with the glides /j/ and /w/ as the endpoints of long monophthongs and diphthongs) are common even in recent American publications. The present chapter uses IPA-based systems (with pure long monophthongs and exclusively vocalic elements in the diphthongs), which have been dominant in British publications since Daniel Jones’s time, i.e. since the early 20th century (e.g. Jones 1909). A conspectus of the current majority view of each ME long vowel’s developmental path is offered in Table 48.2 and Figure 48.1. Table 48.2: Paths from Middle English long vowels to RP pronunciations Middle English (I) (II) (III) (IV) (V) (VI) (VII)
iː uː eː oː ɛː ɔː aː
Modern English (RP) > >
ɪi ʊu
> > >
eː oː æː
i ii/ij
ei/ej
e
υu/υw
əi/əj əυ/əw
(ε) ai/aj aυ/aw
>
ɛː
> > > > > > >
əɪ əʊ
> >
oʊ eː
> >
aɪ aʊ i: u: iː əʊ eɪ
u
o oυ/ow
əυ
(ɔ)
(a) Figure 48.1: Paths from Middle English long vowels to RP pronunciations (Great Vowel Shift and subsequent developments)
Each arrow type (e.g. a sequence of arrows consisting of a dotted line) in Figure 48.1 represents one vowel trajectory, where the arrows with big arrowheads are part of the GVS and those with thin arrowheads are regarded by the majority of researchers as post-GVS developments. As will be shown in some detail in later sections, the changes of the ME
48 Early Modern English: The Great Vowel Shift vowels in the lower half of the phonetic space (the vowels given in brackets) start considerably later than those in the upper half. As was mentioned above, there is broad consensus in recent studies that at least the changes starting in the upper half, i.e. paths (I) to (IV), belong to the GVS or “GVS proper” (Lass 1992, 1999, 2006; Labov 1994: 234; Stockwell 2002; Krug 2003a; McMahon 2006a; see next section for detail). Furthermore, variants like [ɪi/ɪj] and [ʊu/ʊw] of diphthongization stages are not purely notational in Figure 48.1. If Present-day English can serve as a guide, the phonetically most realistic assumption is that both pairs were essentially in complementary distribution: ME [ɪi] and [ʊu] would then be prototypical realizations in prepausal and preconsonantal contexts, while [ɪj] and [ʊw] are prevocalic prototypes serving to avoid hiatus (on the loss of hiatus during Middle English see Section 4 below). Finally, for the paths of ME /iː/ and /uː/ – (I) and (II) in Table 48.2 – some authors have used a more back first element for the modern RP vowels [aʊ] and [aɪ], namely [ɑʊ] and [ɑɪ], respectively, while others again have used intermediate stages [ʌɪ] and [ʌʊ]. Early accounts including Jespersen (1909), Chomsky and Halle (1968), and Wolfe (1972) assume peripheral diphthongization paths for the ME long high vowels, i.e. ME /iː/ via [ei/ɛi] and /uː/ via [ou/ɔu]. Most recent research converges on the central path, one of the reasons being the non-merger of the ME phonemes /iː/ and /ai/ (see Labov 1994: 234 for details). The two alternative paths are, however, to a certain extent compatible if we assume competing standard variants (cf. Lass 1999: 102). In summary, for a number of reasons at least the epithet “great” seems justified for the series of changes under discussion here. Not a single long vowel of the major standard PDE varieties has remained in the position it occupied during the 14th century; the ModE reflexes differ greatly in quality from their ME ancestors and did so at the beginning of the 17th century, to which a number of authors date the end of the GVS (cf. Tables 48.3 and 48.4 for detail); the great majority of modern speakers – including modern speakers of English varieties that descend from dialects which did not participate in all GVS-related changes between 1200 and 1800 – command variants that are somewhere on the paths given in Figure 48.1. And finally, to conclude on a utilitarian or didactic note, about half of the apparent mismatches between modern English orthography and pronunciation are related to the changes sketched in Tables 48.1 and 48.2. Once we have understood the history of the long vowels, such mismatches become more systematic and we can enhance considerably the chances for students of English to deduce the pronunciation from the spelling and vice versa.
2.2 Why a “shift”? Hock (1991: 156) refers to chain shifts as “developments […] in which one change within a given phonological system gives rise to other, related changes.” Generally, two types of shift are distinguished: (i) “drag chain” (or “pull chain”) shifts, which are motivated by the gaps resulting from a vacated space into which other, adjacent phonemes are pulled; (ii) “push chain” shifts, in which one phoneme encroaches onto an adjacent segment’s phonetic space and thus causes the former occupant of this space to shift away (cf. Hock 1991: 156–157; Thomas 2006: 486; for a more detailed discussion of definitional issues involved in shifting, see also the section Metaphors we shift by below). Let us start from a bird’s eye perspective and briefly list aspects that have been advanced in favor of chain shifting from an early date onwards. These usually exploit the notions of symmetry (front vs. back vowels) and gap or slot filling, which explains why
761
762
V Early Modern English the label of “shift” has been particularly attractive to researchers from a structuralist background. Figure 48.1 illustrates the following points: – The high vowels – both front and back – diphthongized, most probably via a central path involving nucleus-glide dissimilation (cf. Section 4 below). – All non-high vowels – both front and back – raised, most probably via a peripheral path. – Diphthongization occurred only for the two highest positions. The question of timing is essential to determining whether the changes are interlinked and whether we are dealing with a “push chain” or a “drag chain” shift. So let us now turn to the question of when the changes shown in Table 48.2 occurred. It is, of course, impossible to pin down exact dates for historical sound changes, inter alia because (a) there exists a gap between writing and speech, (b) progressive pronunciations always coexist with conservative ones, and (c) modern sociolinguistic research into ongoing sound change has revealed that a complex network of social, stylistic, and regional factors plays a role in the distribution of the variants (as well as in the adoption of some of them in the eventual standard). Table 48.3 is a synopsis of previous scholarship (notably Stockwell 1972 and his subsequent work; Lass 1976 and his subsequent work; Faiß 1989; Go¨rlach 1991, 1994; Barber 1997), all of which is essentially based on the interpretation of spelling evidence, rhyming conventions in poetry, dictionaries of rhyming words, as well as early modern English orthoepists’ descriptions, such as those by John Hart (1551) or Alexander Gil (1619). Modern analysts generally assume that the pronunciations featured in Table 48. 3 were common and stylistically unmarked in mainstream southern English speech around the dates given and that progressive dialects anticipate such pronunciations by at least fifty years. Table 48.3: Dating the changes of Middle English long vowels Middle English c.1300 (I) (II) (III) (IV) (V) (VI) (VII)
iː uː eː oː ɛː ɔː aː
c.1500 > > > > >
ɪi ʊu iː uː e̞ ː
c.1600
c.1700
> >
əɪ əʊ
> >
aɪ aʊ
> > >
eː oː æː > ɛː
>
iː
>
eː
Modern English (RP) ≈ ≈ ≈ ≈ ≈ > >
aɪ aʊ iː uː iː oʊ > əʊ eɪ
Table 48.3 suggests an early phase of interrelated changes from about 1300 to 1500 (some authors assume that the changes started about 100 years earlier). A second phase looks likely for the time between 1500 and the second half of the 17th century, when the first merger of two long vowels occurs, which leads to the homophony of see and sea. There can be little doubt that the first phase indeed constitutes a chain shift because the ME pairs /iː/ and /eː/ as well as /uː/ and /oː/ change in lockstep, and in each pair the latter supplants the former. This makes it almost impossible to deny a causal link (be that pushing or dragging). On closer inspection, it becomes clear that we are already dealing with two chain shifts in the upper half because the changes
48 Early Modern English: The Great Vowel Shift in the front and back are merely parallel, not causing each other – except if one wanted to invoke a general upward drift (the great vowel drift?) or a tendency towards parallelism. But it seems rather implausible to conceive of a reason why an upward drift in the back should trigger an upward drift in the front or vice versa. (Keeping this in mind, I will nevertheless, in line with the vast majority of researchers, continue to refer to these two parallel subshifts as one joint shift in order to avoid confusion.) Whether or not we are dealing with one extended shift from 1300 to 1700 or with two independent subshifts in the upper and lower half of the vowel space (as, for instance, Johnston 1992 believes) depends, inter alia, on whether we see the raising of ME /ɛː/ (which would then set off the raising of ME /aː/) and the raising of /ɔː/ as interlinked with, i.e. motivated by the prior raisings from ME /eː/ to [iː] and /oː/ to [uː] respectively. This would be the most encompassing drag chain view of the Great Vowel Shift, where /ɛː/ and /ɔː/ fill the gaps left by the departure of the next higher vowels and ME /aː/ would be dragged into the position of /ɛː/. The question of “Chain shifting or not?” therefore turns out to be definitional rather than factual in nature because the label is legitimate only if we allow time gaps of about 100 years as instances of gap filling (cf. also Guzma´n-Gonza´lez 2003). The issue becomes a bit more complicated because the gap between the two subshifts can be closed if we take raisings by about half a step – ME /ɛː/ to [e̞ ː] and ME /aː/ to [æː] – into consideration.
2.3 Revisiting Dobson, reconsidering shifting In discussions of the GVS two methodologies prevail: one type of analysis concentrates on a limited number of dialects or orthoepists; the alternative approach uses spelling evidence irrespective of the dialect area. Both approaches have their uses: the first is helpful to understand the relationships between different phonemes and the interrelatedness of ongoing changes in a given speech community. The second identifies incipient signs of changes. Both of them are thus valuable for approaching the inception problem of the GVS, although from different angles. At the same time, both approaches pose methodological questions, in particular for the dating of changes. Does early evidence suggest consistent or sporadic allophonic variation in spoken English? Is it representative of a region or just an idiolect? Do medieval and Renaissance writers pay equal attention to all vowels or focus on those that may cause misunderstandings due to their potential for mergers? A related question is whether early sources record raising or lowering of monophthongs earlier than incipient diphthongization, as Dobson (1968: 659) believes. Furthermore, to what extent does spelling or writing about pronunciations reflect the actual pronunciation of a (however specific) speech community and to what extent does it reflect an idealized model? To address such questions, a more global perspective should complement the above approaches. In an attempt to minimize the potential of oversight and overstatement by individuals, this chapter will use Dobson’s (1968) qualitative interpretations in the following quantitative way: each pronunciation in Table 48.4 is listed for a given point in time as soon as at least half of the sources cited in Dobson record a significant change, where “significant change” is defined in a pretheoretical but unbiased manner as “transcribable by IPA vowel symbols without the addition of diacritic marks” (for details, see Krug and Werner 2009). This quantitative method prevents isolated progressive or
763
764
V Early Modern English conservative dialects from entering the present discussion of the GVS. The lines below each pronunciation specify the periods during which the proportion of dialects with the respective pronunciation rises from a third to 50%. In order to avoid spurious findings, such intervals are only given for variants that have at least nine sources in Dobson. Following this method, we can chart cross-dialectal parallel changes and relative time gaps between different vocalic changes. (There exist, of course, some problematic classifications, for instance the categorization of [ʌʊ], [ʌɪ], and [ɛɪ], which are, following Dobson (1968: 660–661), classified in Table 48.4 with [əʊ] and [əɪ], respectively, but could potentially be categorized as further intermediate stages.) What a quantitative survey of Dobson’s sources can serve to do, then, is essentially two things: first, identify in a principled way from what point in time onwards a certain pronunciation is common in many dialects of England; second, identify periods during which many dialects undergo the same (or similar) change. In other words, we can identify when there was a spurt across dialects towards a new pronunciation. This approach has its limitations, too: for one, since Dobson’s sources almost all date from the late 16th and 17th centuries, it goes without saying that this methodology is not appropriate to establish a detailed account of the early phase of the GVS until 1550. It is the period between 1600 and 1650 for which we can draw our conclusions most confidently and which has thus the finest differentiation on the time axis in Table 48.4. Furthermore, this approach cannot serve to identify incipient stages of individual changes. Against this background and in view of the fact that on paths (I) and (II) [əɪ] and [əʊ] include more advanced diphthongizations, it comes as no surprise that all datings in Table 48.4 (except the uncertain ones marked by an asterisk) are later than in Table 48.3, even though the chronologies of Tables 48.3 and 48.4 are surprisingly congruent overall. Time gaps between 10 and 50 years suggest, however, that not all pronunciations of Table 48.3 were actually mainstream as early as hypothesized by previous scholarship. Table 48.4: Majority pronunciations according to Dobson (1968) Middle English c.1300
1500
1550
1600
1620
1630
1640
1650
(I)
iː
ɪi
(II)
uː
ʊu
(III)
eː
iː
iː
(IV)
oː
uː
uː
(V)
ɛː
(VI)
ɔː
(VII)
aː
əɪ (or ɛɪ, ʌɪ)
eː
æː
Late Modern to Modern English (RP) aɪ/ɑɪ
əʊ (or ʌʊ)
aʊ/ɑʊ
iː*
ɛː
*Fewer than five sources in Dobson.
1700
iː
oː
oʊ > əʊ
eː*
eɪ
48 Early Modern English: The Great Vowel Shift Most importantly, Table 48.4 suggests two critical cross-dialectal phases: the first is complete by 1500 and involves the four ME high and mid-high vowels /iː, uː, eː, oː/. This concurs with previous research on the early – and according to some researchers only – stage of the shift. As Dobson contains no earlier sources, 1500 is the terminus ante quem for the first changes affecting ME /iː, uː, eː, oː/. While his sources do not commonly report the early diphthongal pronunciations, Dobson (1968: 659) notes, with reference to Australian and Cockney English (and many other varieties and languages could be invoked), that incipient diphthongization often escapes people’s notice. Such developments can therefore be integrated into a Labovian framework (e.g. Labov 1994: 78), as “changes from below vs. above the level of consciousness”, and their sociolinguistic and prestige-related ramifications (on which see, e.g., Labov 2001: 76–77, 196–197, 509–518) could be investigated in modern dialects. The second major phase suggested by Table 48.4 starts in the late 16th century and ends about 1650. During this period the remaining three ME long monophthongs /ɛː, ɔː, aː/ rise and the nucleus-glide dissimilation of /ɪi/ and /ʊu/ continues. In view of such periods of overlapping change, it seems difficult to dismiss chain shift scenarios. (The stricter definition requiring the preservation of equidistance is dismissed here because this is difficult to apply in changes involving diphthongizations and difficult to put into practice in phonetic analyses.) The reanalysis of Dobson’s sources thus corroborates the major interpretations of Table 48.3, viz. that we can either speak of two phases of a single great shift or of two smaller shifts. Another observation consistent with both Tables 48.3 and 48.4 is that the first candidate for exclusion from the GVS is path (VI), the raising of ME /ɔː/, because here the time gap since the departure of ME /oː/ is biggest (greater than 100 years), while the overlap and likelihood of interrelatedness with other contemporaneous changes is smallest. Finally, the trajectories of ME /aː/ and /ɛː/ seem to suggest a pushing impulse from the lower vowel in the second half of the 17th century. However, there are fewer than five sources for post-1650 [eː] and [iː] in paths (V) and (VII), so no firm conclusions can be drawn from such datings. It might nevertheless seem tempting to posit a third phase starting in the late 17th century for the second-step raising of ME /ɛː/ to [iː] and for the diphthongization of the reflexes of ME /ɔː/ and /aː/ to [oʊ > əʊ] and [eɪ], respectively, but this is not consistent with even a wide definition of chain shifting. The former development is the first merger of two long vowels and, as late 18th and 19th century developments, the latter two seem simply too late to be part of a chain shift that started around 1300. They should therefore not be included in treatments of the shift proper except for didactic purposes, i.e. for tracing historical pronunciations from modern ones.
2.4 Metaphors we shift by: zebras, constellations, dunes, chess, and musical chairs Let us now tackle the problem of chain shifting in a more principled manner by discussing definitional problems. Two criteria are universally advanced in definitions of shifts (cf. Martinet 1952; Stockwell and Minkova 1988; Stockwell 2002; Gordon 2002): (i) the functional credo of preservation of phonemic contrast (i.e. avoidance of mergers) for two or more changing phonemes and
765
766
V Early Modern English (ii) a causal connection between the changes in question, i.e. Change A must have triggered Change B (and so on). These criteria for chain shifting are fulfilled by mainstream southern English if we consider the period from roughly 1400 to 1650 and if the ME diphthongs are excluded from consideration. Lass (1999) has famously labelled a focus on a limited period and a limited number of phonemes and changes a “constellation” or “zebra” fallacy, which implies that linguists see a zebra or constellation because they want to see a particular pattern. And yet, concentration on a limited period is methodologically unproblematic, in fact unavoidable for any discussion of a historical change. Researchers are free in their decision when the highest descriptive or explanatory potential is achieved for their model, and McMahon (2006a: 174) makes a similar point when arguing that “it is hard to see how we can discuss historical patterns at all except insofar as they are the product of hindsight on the part of linguists.” On the basis of Table 48.4, it is the raising of ME /ɔː/ that seems least connected with the remaining developments and might thus be excluded from GVS accounts. We would then have to date the end of the shift to 1640 rather than 1650 – and thus incidentally exclude the second diphthongization stage of ME /uː/, but not that of /iː/, which does not increase the appeal of the account. Both approaches are equally post-hoc, and, in fact, equally justified as doubting that Dobson’s sources allow such precise datings at all. Whether or not exclusion of ME diphthongs is legitimate, however, depends on methodological perspectives which cannot easily be evaluated positively or negatively: it is a reasonable approach for those who want to study the developments of the (system of) ME long vowels only; but it is not a legitimate procedure for those who want to study the systems of and interactions between long ME vowels and diphthongs. Scholars studying mergers, on the other hand, must include former vowel-(semi-)consonant sequences such as may, eight, sty, night, bow, know that merge with vowels (see Stockwell 2002). On the former – let us for convenience call it the “focus-on-long-vowels-only” – approach we find no phonemic mergers until about 1650 to 1700, when ME /ɛː/ and /eː/ merge. (Individual lexical exceptions can be neglected in a discussion of phonological merger.) On the latter approach (which one might term the “focus-on-merger approach”), we find mergers from an early period onwards. Whether the GVS observes the no-merger condition, then, is a matter of perspective and methodology and thus not a matter that can be verified or falsified. This chapter adopts the “focus-on-long-vowels-only” approach, in part because a discussion of ME diphthongs, vowel-glide, and vowel-consonant sequences would increase the complexity to a level that cannot be handled in a handbook. Readers interested in other phonological changes are therefore referred to Schlu¨ter (Chapter 37). Perhaps the focus on ME long monophthongs and their changes over some 250 years can be conceptualized by an alternative metaphor to Lass’s star constellation: dunes (like vowels), although in a steady state of change, can be measured instrumentally and we can take synchronic snapshots of them. It seems legitimate for researchers studying the changing shape and position of dunes (or vowels) to focus on a specific type or selection of dunes, e.g. underwater dunes (or long vowels), those composed of sand (or monophthongs) vs. those composed of gravel (or vowels followed by glides) or those in a specific area (or vowel space). To be sure, the resulting picture will be incomplete, but not necessarily wrong.
48 Early Modern English: The Great Vowel Shift In summary, if one excludes the developments of ME diphthongs from a discussion of the GVS, then the synopsis of Dobson’s interpretations of orthoepistic evidence presented in Table 48.4 is consistent with the classic description of the GVS as a chain shift, during which all long ME non-high vowels raised by one step and the two high vowels diphthongized. This scenario describes fairly accurately the changes from about 1400 to 1650, i.e. very roughly from Chaucer’s to Shakespeare’s time. (The change from ME /aː/ via [æː] to [ɛː] is only an apparent counterexample as the intermediate step [æː] is only half-way between the low and mid-low position.) At the same time, Table 48.4 confirms that not all changes in (I) to (VII) proceeded in lockstep. ME /ɛː/, /ɔː/ and /aː/ started to change much later than /iː/, /uː/, /eː/, and /oː/. This seems to be a good reason for questioning the unitary nature of the shift or for dividing the shift into two phases. On the other hand, the chronological progression of the changes is precisely what some adherents of both push and pull chain scenarios might interpret as supporting evidence for a chain shift. An important definitional problem is that in the literature on chain shifting there is no consensus on lockstep vs. sequentiality. Some authors consider as definitional for shifting, a lockstep movement of different phonemes (e.g. Stockwell 2002), while many general discussions of shifts (like Hock 1991; Bynon et al. 2003; Thomas 2006; Smith 2007: 75) assume the musical chairs analogy, where one change precedes another. The famous Saussurean chess analogy allows for both lockstep and gap interpretations, as recent discussions have thrown into relief its dynamic potential for discussions of language change (Thibault 1997: 96– 98). The problem is aggravated by the fact that even among musical chairs adherents there exists no consensus on how small or big the time gap between two changes may or must be for them to be considered interrelated, a difficulty we already encountered in the interpretation of Table 48.3. In terms of the classic musical chairs analogy, we might ask: how long may it take for a chair (or a gap in the system) to be filled to still qualify as one and the same game? For those theorists who allow a gap of up to 150 years, according to Dobson’s sources, the whole series of changes from (I) to (VII) can be interpreted as forming a unitary Great Vowel Shift – even though, as pointed out above, it would seem preferable to speak of one shift in the back and one in the front since the two are not interrelated. For those who require lockstep or a maximum time gap of 50 years, however, it will be two smaller shifts (affecting the upper half of the ME vowel inventory) followed by another small chain shift raising ME /aː/ and /ɛː/ in the first half of the 17th century plus an individual, but roughly contemporaneous change from ME /ɔː/ to [oː]. Both of these positions are legitimate and neither one is inherently superior from an analytic point of view.
3 On the history of Great Vowel Shift theories In order to improve our understanding of the origin and succession of GVS theories, it is useful to briefly consider their respective intellectual backgrounds. For dominant strands in the philosophy of science – in particular empiricism, positivism, and Darwinism – have had an impact on linguists who have directly or indirectly contributed to the discussion, be they neogrammarians, traditional dialectologists, Prague school and other functionalists, or modern sociolinguists and phoneticians.
767
768
V Early Modern English
3.1 Phonemes, species, and habitats Most of this chapter was written in 2009, which happens to be the year marking the 100th anniversary of Jespersen’s coining of the term “Great Vowel Shift”. The roots of early GVS theories, however, can be traced back further, as the late 19th century had seen a major paradigm shift in the history of scientific thinking: in the middle of that century, Darwin’s evolutionary theory had replaced earlier theories of the evolution of species. In the development of Great Vowel Shift theories, the analogy between biology and language must have seemed particularly appealing because both evolutionary biology and GVS treatments try to describe and explain change (on issues concerning evolutionary sciences and linguistic change, see also Guzma´n-Gonza´lez 2005). Now 2009 also celebrates the bicentenary of Charles Darwin’s birth and at the same time the 150th anniversary of his ground-breaking work On the Origin of Species, which saw three editions within two years and as many as six editions until 1872. Chronological order and parallelism in reasoning suggest strongly that evolutionary thinking had spread from biology to other scholarly domains by the early 20th century, notably to the domain of language and language change. It is probably no coincidence, therefore, that about half a century after Darwin’s (1859) first edition of the Origin of Species, the two most influential push chain and drag chain theories of the GVS were developed by Luick (1896) and Jespersen (1909), respectively. It should be emphasized, however, that this was by no means a new analogy, as venerable linguistic terms like “morphology” illustrate. Nor has this analogizing come to an end since, as can be seen from more recent theories related to evolution and biology as well as mathematical models (like dynamical systems or chaos theory) with applications to both biology and language (cf. McMahon 1994: Chapter 12; Lass 1997: 291–301; Schneider 1997; Croft 2000, 2006; Mufwene 2001, 2008). In modern terms, both push chain and drag chain theories are essentially ecological niche accounts, in which – on the push chain scenario – one species drives a former inhabitant or competitor out of its habitat or – on the drag chain scenario – one species moves into a niche vacated by another species. Such an ecological theory has considerable appeal for sound change theories because of a number of possible analogies: vowels (like species) can be seen as competitors; vowel spaces of adjacent vowels are analogous to habitats; they may overlap and the spaces into which (say, 95% of) vocalic allophones constituting a phoneme fall may shift. After a century of GVS theories, it seems, however, also necessary to reconsider some of the tenets underlying both push and drag theories that have perhaps for too long gone unchallenged. One general difference is that long stressed vowels (unlike species) rarely become extinct. Also, vowels can merge with neighboring vowels – unlike species. The next section will discuss more concrete problems of early theories.
3.2 What’s wrong with the push chain theory? It is in particular Luick’s push-chain theory which has a few serious logical flaws. Although Luick describes adequately a difference between the south (where both ME high vowels diphthongized) on the one hand and what are now conservative northern English and Scottish English dialects on the other (where the back high vowel did not diphthongize), the conclusion that the Great Vowel Shift must have been a push chain seems rash.
48 Early Modern English: The Great Vowel Shift South (and North before fronting of /oː/) iː uː eː oː ɛː ɔː aː
769 North after fronting iː eː øː ɛː aː
uː ←□ ɔː
Figure 48.2: Southern and northern Middle English long vowel inventories according to Lass (1999: 76)
Adherents to push chain scenarios attribute the fact that northern varieties did not diphthongize their back high vowel to a missing back /oː/, which was fronted to /øː/ in northern dialects in the late thirteenth century (Smith 1996: 99–101; Johnston 1997: 69). Consider Luick’s (1896) original formulation, which has a certain ring of circularity to it: [W]enn also mit einem Wort u¯ nur dort diphthongiert wurde, wo o¯ zu u¯ vorru¨ckte, so ergiebt sich vo¨llig zwingend, dass u¯ nur deswegen diphthongiert wurde, weil o¯ zu u¯ vorru¨ckte und es gewissermassen aus seiner Stellung verdra¨ngte. Wir sind also in den Stand gesetzt, eine causale Beziehung zwischen diesen zwei Lautwandlungen sicher festzustellen (Luick 1896: 78; emphasis original). In brief, if u¯ was diphthongized only in regions where o¯ raised to u¯, then it necessarily follows that u¯ was diphthongized only because o¯ raised to u¯ and thus, as it were, pushed it out of its place. We are therefore in a position to firmly establish a causal relationship between these two sound changes [transl. MK].
Lass (1999) summarizes and refines the push chain position as follows: [N]o dialect has done anything to ME /eː/ like what the North did to ME /oː/, i.e. moved it ‘out of position’ before the GVS. And no dialect has consistent undiphthongised ME /iː/. This makes no sense except in the context of a chain shift beginning with the raising of the long mid vowels. A high vowel diphthongises only if the slot below it is filled by a raisable vowel when the shift begins. If the slot below the high vowel is empty (nothing there to push it out of position), there will be no diphthongisation (Lass 1999: 76–77).
Both quotations show that the push chain scenario is explained ex negativo. The argument is that /uː/ did not diphthongize in northern dialects because there was no adjacent vowel /oː/ to push it out of its place. Although this theory seems intuitively plausible and has been described as “beautiful”, the causal link is underdeveloped. For one, the situation was a great deal more complex than Figure 48.2 suggests (see the detailed discussion in Smith 2007: Chapter 6), and northern varieties had in fact developed long /oː/ prior to the GVS as a reflex of Middle English open syllable lengthening (Smith 1996: 99–101). The number of /oː/ words was obviously lower than in dialects that preserve Old English o¯ words like food, which is why scholars who want to save Luick’s theory can with some justification speak of lower pressures in northern dialects. There are more serious problems in the argumentation, however. First, from a strictly logical perspective, the back high vowel space has no explanatory power for what happens in the front vowel space and vice versa. In other words, if diphthongization occurs in the front, this does not entail that it must occur simultaneously in the back, even if this is what we find in
770
V Early Modern English southern Middle English dialects. Second, long high-vowel diphthongization can happen without concomitant raising of the next lower position, as many Present-day English varieties show (see Foulkes and Docherty 1999). Third, there are modern varieties that diphthongize /iː/ much more noticeably than /uː/, which may be rather stable or centralized (cf. modern RP or standard American English). All this suggests that high-vowel diphthongization in the front and back are (a) independent of each other and (b) independent of the existence of a lower pushing vowel. After all, long (or half-long) mid-high vowels exist only in some modern English dialects as allophones of the RP phonemes /eɪ, əʊ, ɔː/ in words like say, so, or force. A last problem for Luick’s and Lass’s push chain theories is that there is no a priori reason why only a mid-high back vowel /oː/ should be able to push /uː/. Although there may be a greater probabilistic likelihood for front vowels to raise along a front path, in principle, any adjacent vowel could have pushed /uː/ out of its position. Fronted northern ME /øː/ could therefore have pushed /uː/ equally well as /oː/, because no long vowel was on the trajectory between /uː/ and /øː/ in the relevant period either. Admittedly, the path from [ø] to [u] is somewhat longer than from [o] to [u], but if we consider the large phonetic space that other vowels travelled during and after the GVS, minor differences in spatial distance do not present a convincing argument for or against certain paths. This is particularly true for /øː/ and /uː/, which are both rounded and thus rather similar from an overall articulatory point of view. In conclusion, if diphthongization of /uː/ does not happen in northern English varieties, the failure of this change to occur cannot be logically linked to the absence (or limited presence) of /oː/. The push chain theory in its current form is therefore to be rejected. Notice that rejecting a causal link between /oː/-fronting and the absence of /uː/-diphthongization in the north does not entail an outright rejection of the push chain scenario. It is in principle possible for /eː/ and /oː/ to have initiated the shift in the south by pushing the higher vowels out of their habitats. But – and this is the last counterargument to Lass’s justification of the push chain scenario – if two adjacent vowels change, it is not necessarily because an adjacent vowel pushes. It may be helpful to invoke the habitat analogy again: species /iː/ may prefer a new habitat for reasons independent of /eː/’s possible occasional inroads into its habitat. Other motivations for /iː/’s move may include a complex of factors like supply of water, food, and sun, all of which would be analogues to phonetic or other motivations for a vowel to change beyond a pushing neighbour. And there may finally be no apparent reasons at all for a vowel to change, not even a pulling neighbour, and yet it does change. What, then is this chapter’s conclusion regarding the inception problem? Lass (1976, 1999) finds no evidence of a clear chronological order, while Stenbrenden (2003) appears to have found evidence of very early high-vowel diphthongization and thus supports the drag chain scenario. The present author also favors the drag chain scenario for the majority of dialects, one reason being uniformitarianism: many modern English dialects diphthongize their high vowels (see the synopsis in Krug 2003a) but have not (or not yet) raised their lower vowels. A second reason is that many northern English and Scottish dialects have followed or are currently following the diphthongization path of /uː/ (see the synopsis in Stuart-Smith 2003). Such dialects can thus be interpreted as conservative rather than as true exceptions to the GVS because adaptation due to contact with southern English as the sole explanation for the diphthongization can be excluded for these varieties on phonetic grounds (see Section 4). In addition, there is a strong historical and crosslinguistic argument against an explanation in terms of contact: there are many related as well as unrelated languages that – at different stages in the past 500 years – underwent
48 Early Modern English: The Great Vowel Shift high-vowel diphthongizations similar to those of the GVS. The contact situations of these languages and of the Middle English dialects that were affected by the GVS, however, are simply too diverse for contact with southern standard English to be considered as the sole or even major explanatory force. The ultimate jury on pushing and pulling may still be out, then, but perhaps such a verdict is not necessary. “English” is not and has never been a monolithic block and it seems quite conceivable that different dialects followed different routes (see, e.g., Knappe 1997 on the development of ME [x] in syllable-coda position). If one adopts this perspective, both the “dialect problem” and the “inception problem” lose some of their poignancy.
4 Motivating the Great Vowel Shift and avenues for further research The question of why the changes known as the GVS happened is not often asked. In other words, accounts of motivation or causation are rare in the literature, unless we include the countless contributions to the inception issue (some of which are summarized in Stockwell and Minkova 1988) and ad-hoc accounts for individual dialects under the rubric of explanations. It is in this area, therefore, that future research seems most promising and new insights can be expected from the digitization of medieval and early modern English texts. Social accounts of causation in the vein of Smith (1996, 2007), who capitalizes on the famous Mopsae argument of hyperadapting incomers (cf. Alexander Gil 1619), are also appealing but difficult to corroborate empirically in the absence of unambiguous historical sociolinguistic evidence or modern parallel cases. As long as there are no detailed sociophonetic accounts, the most realistic path to a motivation theory would be one that appeals to more general principles of phonetic and phonological change. What comes closest to such a crosslinguistic motivation are two of the recurrent tendencies identified for chain shifting and granted principle status by Labov (1994: 116, 176): Principle I. In chain shifts, long vowels rise. Principle IIa. In chain shifts, the nuclei of upgliding diphthongs fall. Principle I can accommodate the raisings of all non-high vowels in the GVS, while Principle IIa captures the diphthongization paths (I) and (II) of Table 48.2, which are for convenience repeated below with the minimal addition of a moraic representation for the ME long vowel starting points: > ɪi
> əɪ > aɪ
(I)
iː = ii
(II)
uː = uu > ʊu > əʊ > aʊ
An alternative (but compatible) phonetically driven approach is the optimalitytheoretic account by Minkova and Stockwell (2003), who focus on the nucleus-glide dissimilation of the same diphthongizations, i.e. the increasing phonetic distance between the first and second element of these diphthongs. They argue convincingly that this process creates more optimal diphthongs from a hearer’s perspective because the likelihood of misunderstanding decreases.
771
772
V Early Modern English If we subscribe to the drag chain scenario, then a hearer-based economy can be invoked for the subsequent filling of the high-vowel spaces, too. This follows from the functionalist principle of maximal differentiation, which was formulated and refined by Martinet (e.g. 1952) but had implicitly been utilized by historical linguists arguing for gap filling since at least the 19th century, including the GVS chain shift advocates from both camps. According to this principle, it is useful for languages to have the extreme positions /a, u, i/ filled to maximize the distance between the distinctive vowels in the available vowel space, and indeed there are very few languages that lack one of these three vowels (Ladefoged and Maddieson 1996). Researchers therefore speak of an “unbalanced system” when the two high vowel positions are empty and assume that they are likely to be refilled soon. Language is the constant negotiation between hearer-based and speaker-based economies, so it would be surprising if speaker-based principles did not play a role in the GVS. Elsewhere (Krug 2003a), I have presented arguments in terms of speaker economy pointing in a similar direction as the principles and optimality-theoretic accounts cited above, thus strengthening the case for the drag chain scenario. The arguments presented involve phonetic factors that exploit the tense-lax opposition, hiatus avoidance, and the sonority hierarchy with its implications for high-vowel diphthongization. In essence, I argue that the instability of long high vowels is due to their relatively high production effort: since high vowels are more tense than low vowels and since pure [i] and [u] are more peripheral, their production (in particular when they are long) involves more muscular effort than that of lower vowels. Long high vowels are therefore assumed to be intrinsically prone to diphthongization, which is well supported not only by English but also by crosslinguistic evidence (Wolfe 1972: 131–134; Krug 2003a). The first stages [ɪi, əɪ] and [ʊu, əʊ] in high-vowel diphthongization along a central path are interpreted as lenition that is led by high frequency items, notably pronouns like thou, I, my, thy. A similar case for lenition has been made by Feagin (1994) for the monophthongization of /aɪ/ in southern American English, which seems to be led by the pronouns I and my. Such high-frequency items tend to develop progressive variants below the level of consciousness (Krug 2003b), which may be the impulse for a shift of a phoneme’s prototypical realization and thus of its positional displacement. An additional argument for early diphthongization in terms of speaker economy derives from the fact that the loss of epenthetic [ʔ] in hiatus contexts (on which see Minkova 2003) is roughly contemporaneous with the beginning of the GVS. From a usage-based perspective, then, it seems likely for the two most frequent pronounverb sequences of English (I-am and thu-art) to develop intrusive glides (/j/ and /w/ respectively) at the former word boundaries. The matter is more complex for I-am due to the history of the first person pronoun, but there is clear evidence for an increase of potential hiatus contexts from the historical Helsinki Corpus (on the development from ME [ic¸] via [ij] to [iː] see Dobson 1968: 667). Even if, as seems likely, glottal onset before potential low-stress items like am and art or between tightly bonded sequences like thu-art was infrequent or did not exist in early ME at all, the liaison argument remains nevertheless valid: the development of intrusive glides would merely have to be antedated. In any case, the resulting pronunciations of the pronouns I and th(o)u in these high-frequency sequences would have resembled open-syllable diphthongization of words like my, thy and thou in isolation. On that view, two independent
48 Early Modern English: The Great Vowel Shift phonetic, hence natural tendencies of high-vowel diphthongization mutually reinforced each other. And yet, high-vowel diphthongization may not be the full answer to the issues of inception and causation. Modern phonetic research (see for instance the gamut of studies presented in Labov 1994: Chapters 6 and 8) allows for simultaneous change as it suggests that the reality is neither fully congruent with lockstep movement nor with a major time gap: synchronically, vowel spaces of adjacent phonemes overlap, especially so during ongoing change, where one phoneme encroaches on the space of an adjacent phoneme. This situation holds for a single speaker, is common within any speech community and normal for different dialects. Detailed quantitative phonetic and sociolinguistic research of conservative and progressive speech communities, e.g. northern England, Scotland, Australia, New Zealand, or London could therefore throw new light on the historical GVS. Another area that deserves more attention in future research (and not only on GVSrelated research) is the role of allophonic variation of vowels, the abundance of which has led some researchers to reject the existence of phonemes altogether (see e.g. Kretzschmar and Tamasi 2003). Without a doubt, more research is necessary on the effects of high-frequency items and sequences (cf. the studies in Bybee and Hopper 2001) as well as of syllable type (e.g. open vs. closed) and neighboring sounds in such sequences as me/my bike or It was me. And yet, it is almost surprising how regular and parallel the changes were that affected the allophones of each ME long vowel and such regularity points indeed to the cognitive reality of more abstract, phonemic representations. In conclusion, I still tend to believe, as in 2003, that the most likely answer to the question of who triggered the GVS is: “You and me, basically; and maybe also he and she, or us and we. All of us essentially.” But a lot more detailed socio-phonetic research and theoretical refinement will be necessary before we can turn this hypothesis into yet another theory that students of English historical linguistics should consider for memorization. Students might consider, however, discussing the many GVS-related hypotheses and debates mentioned in this chapter as heuristics for critically evaluating and better understanding the nature of linguistic change and theory building.
5 References Barber, Charles. 1997. Early Modern English. Cambridge: Cambridge University Press. Bloch, Bernard and George L. Trager. 1942. Outline of Linguistic Analysis. Baltimore: Linguistic Society of America. Bybee, Joan. 2003. Mechanisms of change in grammaticization: The role of frequency. In: Brian D. Joseph and Richard D. Janda (eds.), The Handbook of Historical Linguistics, 602–623. Oxford: Blackwell. Bybee, Joan and Paul Hopper (eds.). 2001. Frequency and the Emergence of Linguistic Structure. Amsterdam/Philadelphia: John Benjamins. Bynon, Theodora, Elizabeth Closs Traugott, Anthony Kroch, Rex E. Wallace, and Henry M. Hoenigswald. 2003. Language change. In: William J. Frawley (editor-in-chief), International Encyclopedia of Linguistics. 2nd edn. Oxford: Oxford University Press. (e-reference edition, accessed 11 November 2009). Chomsky, Noam and Morris Halle. 1968. The Sound Pattern of English. New York: Harper and Row.
773
774
V Early Modern English Christy, T. Craig. 1983. Uniformitarianism in Linguistics. Amsterdam/Philadelphia: John Benjamins. Croft, William. 2000. Explaining Language Change: An Evolutionary Approach. Harlow: Longman. Croft, William. 2006. Evolutionary models and functional-typological theories of language change. In: van Kemenade and Los (eds.), 68–91. Darwin, Charles. 1859. On the Origin of Species by Means of Natural Selection. London: Murray. Dobson, Eric J. 1968. English Pronunciation 1500–1700. 2 vols. Oxford: Oxford University Press. Faiß, Klaus. 1989. Englische Sprachgeschichte. Tu¨bingen: Franke. Feagin, Crawford. 1994. “Long I” as a microcosm of southern states speech. Paper given at NWAV1994. Ms., Zurich University. Foulkes, Paul and Gerard Docherty (eds.). 1999. Urban Voices: Accent Studies in the British Isles. London: Arnold. Gil, Alexander. 1619. Longonomia Anglica. (2nd ed., 1621). London: Beale. Gordon, Matthew J. 2002. Investigating chain shifts and mergers. In: J. K. Chambers, Peter Trudgill, and Natalie Schilling-Estes (eds.), The Handbook of Language Variation and Change, 244–266. Oxford: Blackwell. Go¨rlach, Manfred. 1991. Introduction to Early Modern English. Cambridge: Cambridge University Press. Go¨rlach, Manfred. 1994. Einfu¨hrung ins Fru¨hneuenglische. Heidelberg: Winter. Guzma´n-Gonza´lez, Trinidad. 2003. Revisiting the revisited: Could we survive without the Great Vowel Shift? Studia Anglica Posnaniensia 39: 121–131. Guzma´n-Gonza´lez, Trinidad. 2005. Out of the past: A walk with labels and concepts, raiders of the lost evidence and a vindication of the role of writing. International Journal of English Studies 5: 13–32. Hart, John. 1551. The opening of the unreasonable writing of our inglish toung. Unpublished Ms. Hock, Hans Heinrich. 1991. Principles of Historical Linguistics. 2nd edn. Berlin/New York: Mouton de Gruyter. Jespersen, Otto. 1909. A Modern English Grammar on Historical Principles. Vol. I Sounds and Spellings. London: Allen and Unwin. Johnston, Paul A. 1992. English vowel shifting: One great vowel shift or two small vowel shifts? Diachronica 9: 189–226. Johnston, Paul A. 1997. Older Scots phonology and its regional variation. In: Charles Jones (ed.), The Edinburgh History of the Scots Language, 47–111. Edinburgh: Edinburgh University Press. Jones, Daniel. 1909. The Pronunciation of English. Cambridge: Cambridge University Press. Kemenade, Ans van and Bettelou Los (eds.). 2006. The Handbook of the History of English. Oxford: Blackwell. Kretzschmar, William A., Jr., and Susan Tamasi. 2003. Distributional foundations for a theory of language change. World Englishes 22: 377–401. Knappe, Gabriele. 1997. Though it is tough: On regional differences in the development and substitution of the Middle English voiceless velar fricative [x] in syllable coda position. (Special issue on Language in Time and Space: Studies in Honour of Wolfgang Viereck on the Occasion of his 60th Birthday, ed. by Heinrich Ramisch and Kenneth Wynne.) Zeitschrift fu¨r Dialektologie und Linguistik 97: 139–163. Krug, Manfred. 2003a. (Great) vowel shifts present and past: Meeting ground for structural and natural phonologists. Penn Working Papers in Linguistics (Selected papers from NWAVE 31 at Stanford) 9(2): 107–122. Krug, Manfred. 2003b. Frequency as a determinant in grammatical variation and change. In: Gu¨nter Rohdenburg and Britta Mondorf (eds.), Determinants of Grammatical Variation in English, 7–67. Berlin/New York: Mouton de Gruyter. Krug, Manfred and Valentin Werner. 2009. Dobson revisited: Long vowel variation and the Great Vowel Shift. Ms. Bamberg. Labov, William. 1994. Principles of Linguistic Change. I: Internal Factors. Oxford: Blackwell.
48 Early Modern English: The Great Vowel Shift Labov, William. 2001. Principles of Linguistic Change. II: Social Factors. Oxford: Blackwell. Ladefoged, Peter and Ian Maddieson. 1996. The Sounds of the World’s Languages. Oxford: Blackwell. Lass, Roger. 1976. Rules, metarules and the shape of the Great Vowel Shift. In: Roger Lass (ed.), English Phonology and Phonological Theory: Synchronic and Diachronic Studies, 51–102. Cambridge: Cambridge University Press. Lass, Roger. 1992. Phonology and morphology. In: Norman Blake (ed.), The Cambridge History of the English Language. Vol. II. 1066–1476, 23–155. Cambridge: Cambridge University Press. Lass, Roger. 1997. Historical Linguistics and Language Change. Cambridge: Cambridge University Press. Lass, Roger. 1999. Phonology and morphology. In: Roger Lass (ed.), The Cambridge History of the English Language. Vol. III. 1476–1776, 56–186. Cambridge: Cambridge University Press. Lass, Roger. 2006. Phonology and morphology. In: Richard Hogg and David Denison (eds.), A History of the English Language, 43–109. Cambridge: Cambridge University Press. Luick, Karl. 1896. Untersuchungen zur englischen Lautgeschichte. Strassburg: Truebner. Martinet, Andre´. 1952. Function, structure and sound change. Word 8(1): 1–32. McMahon, April. 1994. Understanding Language Change. Cambridge: Cambridge University Press. McMahon, April. 2006a. Restructuring Renaissance English. In: Lynda Mugglestone (ed.), The Oxford History of English, 147–177. Oxford: Oxford University Press. McMahon, April. 2006b. Change for the better? Optimality theory vs. history. In: van Kemenade and Los (eds.), 3–23. McMahon, April. 2007. Who’s afraid of the vowel shift rule? Language Sciences (Issues in English phonology) 29(2–3): 341–359. Minkova, Donka. 2003. Alliteration and Sound Change in Early English. Cambridge: Cambridge University Press. Minkova, Donka and Robert Stockwell. 1999. Explanations of sound change: Contradictions between dialect data and theories of chain shifting. In: Clive Upton and Katie Wales (eds.), Dialectal Variation in English: Proceedings of the Harold Orton Centenary Conference 1998, 83–102. Leeds: School of English. Minkova, Donka and Robert Stockwell. 2003. English vowel shifts and “optimal“ diphthongs: Is there a logical link? In: D. Eric Holt (ed.), Optimality Theory and Language Change, 169–190. Amsterdam: Kluwer. Mufwene, S. Salikoko. 2001. The Ecology of Language Evolution. Cambridge: Cambridge University Press. Mufwene, S. Salikoko. 2008. Language Evolution: Contact, Competition and Change. London: Continuum. Ogura, Mieko. 1987. Historical English Phonology: A Lexical Perspective. Tokyo: Kenkyusha. Schneider, Edgar. 1997. Chaos theory as a model for dialect variability and change? In: Alan A. Thomas (ed.), Issues and Methods in Dialectology, 22–36. Bangor: University of Wales, Department of Linguistics. Schu¨tzler, Ole. 2009. Unstable close-mid vowels in Modern Scottish English. In: Carlos PradoAlonso, Lidia Go´mez-Garcı´a, Iria Pastor-Go´mez, and David Tizo´n-Couto (eds.), New Trends and Methodologies in Applied English Language Research: Diachronic, Diatopic and Contrastive Studies, 153–182. Bern: Peter Lang. Smith, Jeremy. 1996. An Historical Study of English: Function, Form and Change. London: Routledge. Smith, Jeremy. 2007. Sound Change and the History of English. Oxford: Oxford University Press. Stenbrenden, Gjertrud F. 2003. On the interpretation of early evidence for ME vowel-change. In: Barry J. Blake and Kate Burridge (eds.), Historical Linguistics 2001: Selected papers from the 15th International Conference on Historical Linguistics, Melbourne, 13–17 August 2001, 403– 415. Amsterdam/Philadelphia: John Benjamins.
775
776
V Early Modern English Stockwell, Robert. 1972. Problems in the interpretation of the Great Vowel Shift. In: M. Estellie Smith (ed.), Studies in Linguistics in Honor of George L. Trager, 344–362. The Hague: Mouton. Stockwell, Robert. 2002. How much shifting actually occurred in the historical English vowel shift? In: Donka Minkova and Robert Stockwell (eds.), Studies in the History of the English Language: A Millennial Perspective, 267–281. Berlin/New York: Mouton de Gruyter. Stockwell, Robert and Donka Minkova. 1988. The English Vowel Shift: Problems of coherence and explanation. In: Dieter Kastovsky, Gero Bauer, and Jacek Fisiak (eds.), Luick Revisited, 355–394. Tu¨bingen: Gunter Narr. Stuart-Smith, Jane. 2003. The phonology of Modern Urban Scots. In: John Corbett, J. Derrick McClure, and Jane Stuart-Smith (eds.), The Edinburgh Companion to Scots, 110–137. Edinburgh: Edinburgh University Press. Thibault, Paul J. 1997. Re-reading Saussure: The Dynamics of Signs in Social Life. New York: Routledge. Thomas, E. R. 2006. Vowel shifts and mergers. In: Keith Brown (ed.), Encyclopedia of Language and Linguistics, 484–494. 2nd edn. Amsterdam: Elsevier. Trager, George L. and Henry Lee Smith, Jr. 1951. An Outline of English Structure. Studies in Linguistics, Occasional Papers 3. Norman, OK: Battenburg Press. Wolfe, Patricia M. 1972. Linguistic Change and the Great Vowel Shift in English. Berkeley: University of California Press.
Manfred Krug, Bamberg (Germany)
49 Early Modern English: Relativization 1 2 3 4 5 6 7
Introduction Written and speech-related Early Modern English: that, wh-forms and zero The relative clause: that, wh-forms and zero Personal versus nonpersonal antecedents Forms and functions of the relativizers Summary References
Abstract This chapter describes the use of relativizers in Early Modern English, focusing on speech-related material, i.e. Trials and Drama texts from the periods 1560–1599 and 1680–1719. What is most striking as regards the use of relativizers is the predominance of the relativizer that. It is used with all types of personal and nonpersonal antecedents and in restrictive and nonrestrictive relative clauses, although it is rare in nonrestrictive relative clauses. The wh-forms compete with that in restrictive clauses but who is not frequent except with personal names, and which decreases in frequency, as it is rarely used with personal antecedents. The zero construction becomes increasingly frequent, particularly in the second period (1680–1719). Thus, the variation is rather between that and the zero construction than between that and the wh-forms. By 1719, we recognize many of the features of Present-day usage of relativizers. These features probably appeared earlier in speech-related data than in the literary language. Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 776–790
49 Early Modern English: Relativization
1 Introduction In the Early Modern English period (c.1500–1700), the use of relativizers and relative clauses is in many ways different from Present-day English usage. In Early Modern English, the wh-forms increase in frequency, particularly in restrictive relative clauses, especially when the personal (who) / nonpersonal (which) contrast becomes more explicit. Throughout the period, however, that is the most frequent relativizer. Moreover, individual letter writers, scientists, and educators have preferences as regards the use of wh-forms versus that. Previous studies (e.g. Ryde´n 1966; Dekeyser 1984) focus on the use of relativizers in EModE literary and scholarly language whereas this paper will focus on the use of relativizers in spoken language and compare the two uses. “Adnominal relative clauses” (i.e. relative clauses as postmodifiers, see examples 1–3), will be studied to examine the frequency of the relativization strategies that, wh-forms and the zero construction. “Sentential relative clauses” will also be included in the study (see example 4). Here, the whole preceding clause or the predicate or parts of it function as antecedent (see Quirk et al. 1985: 1118–1120). Sentential relative clauses are typical of spoken language (Biber 1988: 106–107) and comment on what has been said before. “Nominal relative clauses”, as in example (5), will not be examined in this paper. (1)
Nightingale. […] I enquired of another who lived in the Mews, if he knew Mrs. Baynton (1702 Trials, Haagen Swendsen)
(2)
S. G. Pray is she a Country Lady that has got a good Joynture. (1702 Trials, Haagen Swendsen)
(3)
Swen. […] I beg it, my Lord, for ’tis the most material thing Ø I have to ask, (1702 Trials, Haagen Swendsen)
(4)
Att. Gen. […] but by proof he did not come in till Twelve, which was after the thing was done. (1680 Trials, John Giles)
(5)
Men. Because she did as you do now, on whom soeuer she met withall, she railed, and therfore well deserued that dogged name. (1595 Drama, Warner, Menaecmi)
The paper will concentrate on the variation between that, wh-forms and zero as regards type of relative clause, antecedent, and the function of the relativizer in the relative clause. Section 2 will present the speech-related material and frequencies of that, wh-forms, and zero in my study, as well as in previous studies on EModE relativization.
2 Written and speech-related Early Modern English: that, whforms and zero The closest representatives of spoken EModE are speech-related texts, such as Trial proceedings and Drama comedy (see Culpeper and Kyto¨ 2000: 186–193). The speech-related texts (henceforth spEModE) in my study are drawn from A Corpus of English Dialogues 1560–1760 (CED) (see Kyto¨ and Culpeper 2006; Kyto¨ and Walker 2006). The Trials include 56,710 words and the Drama texts 87,460. The periods studied are 1560–1599 and 1680–1719.
777
778
V Early Modern English In Table 49.1, the distribution of that, wh-forms, and zero is given as presented in the studies of Ryde´n (1966), Dekeyser (1984), and in the two spEModE periods. To show the distribution of relativizers within each period, the two spEModE periods are treated separately in Table 49.1, but are otherwise throughout the paper treated mainly as one category representing spEModE. Ryde´n’s study describes the use of relativizers in literary and scholarly language as represented by Thomas Elyot (1490?–1546) and his contemporaries and covers the period 1520–1560. Although Ryde´n (1966) concentrates on the early 16th century, most of the results and comments presented there can be taken as representative of the remainder of the EModE period. Ryde´n’s detailed study is thus the main source of written EModE but naturally the time period (i.e. the early 16th century) will be taken into account when comparison is made with the later spEModE texts studied in this paper. Dekeyser’s corpus covers the period 1600–1649 and includes literary texts, prose, and poetry, but also letters and speech-related Drama texts. In Table 49.1, Dekeyser’s Drama texts have been separated from prose, poetry, and letters (wrEModE: 1600–1649) to compare with the distribution of the relativizers in the spEModE material in this paper. Table 49.1: That, wh- forms, and zero in EModE: written and speech-related texts That/wh-form/zero
That
Wh-form
Zero
wrEModE: 1520–1560 (Ryde´n 1966) wrEModE: 1600–1649 (Dekeyser 1984) EModE Drama: 1600–1649 (Dekeyser 1984) spEModE (Trials and Drama): 1560–1599 spEModE (Trials and Drama): 1680–1719
46% 28% 46% 51% 47%
52% 65% 30% 38% 32%
2% 7% 24% 11% 21%
As appears from Table 49.1, that is the most frequent relativizer, particularly in the two periods of spEModE analysed in the study, 1560–1599 and 1680–1719 (47%–51%), and in Dekeyser’s Drama texts from 1600–1649 (46%). That is more common than the whforms taken together. In the Trials studied in this paper, containing recorded authentic dialogue, that is the preferred relativizer, both with members of the legal profession and with the defendants and witnesses. Ball (1996: 246–248), who includes a study of EModE Trials (the State Trials from 1680, 1692 and 1693, and the Salem Witchcraft Trials from 1692), reports the same result: that is the most frequent relativizer whatever speaker role (i.e. members of the legal profession and defendants and witnesses). In the Drama texts included in this paper, that is predominant irrespective of social class. Characters representing the lower social classes could be expected to use that and zero even more frequently (see Barber 1997: 235–236). However, the playwrights use other means, such as spelling or vocabulary, to mirror in writing features of a dialect or of uneducated speech (cf. Culpeper 2001: 206, 209, 212). Dekeyser does not discuss the use of relativizers in terms of social class in his Drama texts from 1600–1649. The wh-forms increase in frequency during the EModE period, but this is only true of the written data: 52% of wh-forms occur in the period 1520–1560 (Ryde´n’s study) and 65% in 1600–1649 (Dekeyser’s study). In spEModE, only 32%–38% of the relative clauses are headed by a wh-form. In Dekeyser’s Drama texts, the frequency of wh-forms is also low: 30% (see Table 49.1).
49 Early Modern English: Relativization
779
The most striking difference between the first and the second spEModE periods in my study is the frequency of the zero construction: it is nearly twice as frequent between 1680–1719 compared with 1560–1599 (21% versus 11%, see Table 49.1). Although the use of the zero construction is not necessarily an indication of informal style in EModE (Rissanen 1984: 430; Visser 1963: 853), a zero relativizer is apparently more typical of spEModE dialogue than of literary and formal language. Zero constructions are also frequent (24%) in Dekeyser’s Drama texts. Ball (1996: 248) notes in her Trials study that zero relativizers are mainly used by speakers other than lawyers and aristocrats (see also Section 4.4). The next section will take a closer look at the relative clause in EModE; more precisely, how the two types of relative clause, restrictive and nonrestrictive, influence the use and frequency of relativizers in spEModE.
3 The relative clause: that, wh-forms and zero Relative clauses are the most frequent type of subordinate clause in EModE literary and scholarly language (Ryde´n 1966: xix, 362). Relative clauses are also common in spEModE (876/100,000 words). Restrictive relative clauses restrict the reference of the antecedent, in terms of classification or identification, whereas a nonrestrictive relative clause merely gives additional information about a previously identified antecedent (cf. Quirk et al. 1985: 1239–1242; Poutsma 1926–1929: 421–430; Ryde´n 1974). A frequent type of nonrestrictive clause in my spEModE data is the sentential relative clause (see Section 1). Table 49.2 shows the frequencies of that, the wh-forms and the zero relativizer in restrictive and nonrestrictive relative clauses in my spEModE material. Restrictive relative clauses are predominant in the spEModE data: 78% of the relative clauses are restrictive. Given their classifying and identifying function, restrictive relative clauses are particularly frequent in the Trials where establishing the identity of people or evidence is crucial. As in Present-day English, the zero construction appears only in restrictive relative clauses. The distribution of the relativizers in restrictive and nonrestrictive relative clauses will be discussed in the following sections. Table 49.2: That, wh- forms, and zero in restrictive and nonrestrictive relative clauses in spEModE That/wh-form/zero
Restrictive
That Wh-forms (total) Who Whom Whose Which Zero
532 188 42 16 30 100 193
Total
913 (78%)
(58%) (21%) (5%) (2%) (3%) (11%) (21%)
Nonrestrictive
Total that/wh-form/zero
(16%) (84%) (18%) (6%) (6%) (53%) —
572 405 88 33 47 237 193
40 217 46 17 17 137
257 (22%)
1170 (100%)
3.1 Restrictive relative clauses Most of the restrictive relative clauses are headed by that, both in written EModE and speech-related data (spEModE: 58%, see Table 49.2). The conjunction-like
780
V Early Modern English character of that signals, in principle, a tighter link to the antecedent and antecedent clause than a relative clause with a wh-form (Ryde´n 1974; Rissanen 1984: 420–422). This is particularly apparent when a pronoun is the antecedent, as in example (6). (6)
Bridges. You said thus, That the Papists were the best Religion, and that those that were not of that Religion were Damn’d. (1680 Trials, John Giles)
In spEModE, who and which are not very frequent in restrictive relative clauses (who: 5% and which: 11%, see Table 49.2) and do not compete with that to any great extent. There is no increase in the use of who and which (examples 7 and 8) during the second spEModE period (1680–1719). (7)
Mr. Att. Gen. Then, my Lord, we have another piece of Evidence which we wou’d offer to your Lordship, which is not direct Evidence against the Prisoner, (1696 Trials, Ambrose Rookwood)
(8)
Wild. […] If a Woman be very handsome, and meets with a Man who has Wit enough to know and value it; the Consequence speaks it self, and needs no Corroborating Evidence. (1696 Drama, Manley, The Lost Lover)
The wh-forms (who, whose, whom, and which) as a whole are not more frequent than the other relativization strategy in restrictive relative clauses, the zero construction. A wh-form or zero occurs in 21% of the cases (see Table 49.2).
3.2 Nonrestrictive relative clauses In nonrestrictive relative clauses, the wh-forms predominate in spEModE (84%, see Table 49.2). Who is fairly common in nonrestrictive relative clauses (18%). Also the inflected forms whom and whose are more frequent in nonrestrictive relative clauses than in restrictive clauses (6%, see Table 49.2), but are still rare. However, the most frequent relativizer in nonrestrictive relative clauses in spEModE is which. In Drama and Trials, more than 50% of the nonrestrictive relative clauses are headed by which (see Table 49.2). As mentioned earlier, sentential relative clauses are typical of speech and speech-related language and they represent a frequent type of nonrestrictive relative clause in my spEModE texts. This influences the frequency of which, since it is the only possible relativizer in such clauses, see example (9). (9)
Mr. Slater. […] I took him for a Frenchman, he used very much to espouse the Interest of the King of France, which I used to chide him for. (1716 Trials, Francis Francia)
Nonrestrictive that is infrequent in the EModE period; both in the early 16th century and in later periods. Ryde´n (1966: 278) reports that in Elyot and his contemporaries’ works, only 10% of the nonrestrictive clauses are introduced by that. In Dekeyser’s (1984) data, which includes prose, poetry, and letters but also speech-related drama
49 Early Modern English: Relativization texts from the early 17th century, only 20% of the nonrestrictive clauses are headed by that. In the spEModE data in this study, nonrestrictive that occurs in 16% of the relative clauses, but it decreases towards the end of the second spEModE period studied (1680–1719). Most of the examples of nonrestrictive that are to be found in the early comedies, see example (10). (10) Hephest. […] You (Alexander) that would be a God, shewe your self in this worse then a man, (1584 Drama, Lyly, Alexander and Campaspe)
3.3 Structural complexity of EModE relative clauses In EModE, the relative clauses can be of considerable structural and semantic complexity, including different types of co-ordinate relative clauses (cf. Ryde´ n 1970: 13–15, 20–27). In spEModE, co-ordinate relative clauses contribute to the complexity, as do sentential relative clauses and relative clauses which expand on the premodifiers and postmodifiers of the noun functioning as antecedent. In example (11), the first of the three relative clauses is sentential. In a sentential relative clause, an argument is summed up or commented on (which was done and which I did occur frequently in the Trials). The other two relative clauses in example (11) are coordinated, including a wh-clause (who) + a clause introduced by zero. Who refers to “Gutbert”. (11) Sollic. This is not so, as hath well appeared. Besides this, the conveying away of Gutbert, which was done by your means, and who decyphered this Letter, and could have disclosed the Matter, proveth a great Guiltiness in you. (1571 Trials, Thomas Howard Duke of Norfolk) In example (12), the relative clause, which is nonrestrictive, expands on the information given in the premodification and completes the characterization of an identified antecedent, a sorry ignorant Knave (see Ryde´n 1974, 1984; Jacobsson 1994). Such relative clauses are frequent in Drama, as shown in example (12). (12) Oliv. Are you so silly to believe it, he seems to be a sorry ignorant Knave, that has more Will than Power to do Evil. (1696 Drama, Manley, The Lost Lover)
4 Personal versus nonpersonal antecedents The personal/nonpersonal contrast becomes increasingly explicit during the EModE period and the most important change is apparently the use of which. Although which does occur with personal nouns, it becomes more restricted to nonpersonal antecedents. Deities are a special group of antecedents in EModE texts and have a very high personal status. It is with this type of antecedent that who starts being used (cf. Ryde´n 1983). As antecedents, deities are more common in written EModE data, not only in religious writing and the Bible but also in letters (Nevalainen and Ramoulin-Brunberg 2002: 118). In spEModE, which, as in example (13), and that are used with deities but the form who is not found. On the other hand, with preposition + relativizer, whom is used.
781
782
V Early Modern English (13) Perin. […] The God of trueth and perfect equitie, Which will reuenge wrong to the innocent, with thousand plagues and tortors worse than death. (1594 Drama, Anon., A Knacke to Know a Knave) Table 49.3 shows the distribution of the relativizers with personal and nonpersonal antecedents in spEModE and how the subject and (direct + prepositional) object functions are realized as that, a wh-form or as zero with the two types of antecedent. The functions of the relativizers will be discussed in more detail in Section 5. Table 49.3: Antecedent and function of that, wh-forms and zero in spEModE Function
Subject
Antecedent
Personal
Non-personal
That Who Whom Which Zero
245 86 4 8 12
196 (62%) 4 (1%)
Total
355
(69%) (24%) (1%) (2%) (3%)
Total Subject
Object Personal
Non-personal
Total Object
Total
18 (25%)
98 (31%)
116
109 (34%) 9 (3%)
441 90 4 117 21
30 (41%) 2 (3%) 23 (31%)
106 (33%) 113 (36%)
30 108 136
557 90 34 225 157
318
673
73
317
390
1063
4.1 That In spEModE, the relativizer that is used as frequently with a personal as with a nonpersonal antecedent, particularly when that functions as subject in the relative clause (see Table 49.3: 62%–69%). Although that is frequent with most types of personal and nonpersonal antecedents, there are certain types of antecedent in spEModE with which it is particularly common. That is often used with all (as pronoun and determiner) and with pronouns such as everything, something, any thing, one, any body, with general nouns such as person(s), people, and thing(s) and with a superlative + noun, as in examples (14)–(16) (cf. also Johansson 2006: 154–166). With many of these types of antecedent, we also find the zero construction when functioning as direct object or as prepositional object (cf. Present-day English, and see Section 4.4). (14) Ld. Townshend. All that Mr. Walpole brought me were laid there, and I saw him take them back again. (1716 Trials, Francis Francia) (15) Smy. […] And, let me tell you, Mr. Wildman, I Love my Wife, and don’t like People that slight her Charms, (1696 Drama, Manley, The Lost Lover) (16) Boy. Go thy waies for the prowdest harlotrie that euer came in our house. (1599 Drama, Chapman, An Humerous Dayes Mirth) Not only indefinite pronouns (e.g. any body, one, some) occur as personal antecedents of that but also personal pronouns and demonstratives (he, you, they, those, this, and who in the combination who that). In example (17), he that is generic (= ‘anyone who’; see Jespersen 1927: 98–99, 120, 154–155).
49 Early Modern English: Relativization (17) Hone. I, but he remembers not where Christ saith, hee that giveth a cup of water in my name shall be blessed. […] He that giueth to the poore lendeth vnto the Lord, (1594 Drama, Anon., A Knacke to Know a Knave) That is used repeatedly in “cleft sentences” in spEModE, particularly in the Trials. Cleft sentences serve the purpose of emphasizing a person’s identity, actions or whereabouts, or the time of a crime, for instance in the statements of witnesses. In cleft sentences (itclefts, th-clefts and wh-clefts), that is used as frequently with personal and as with nonpersonal antecedents. Examples (18)–(19) are from the Trial of Haagen Swendsen, where cleft sentences are particularly common. (18) Judge Powel. How long was it from the time that you were parted after Arresting, that you saw Mrs Rawlins again. (1702 Trials, Haagen Swendsen) (19) Wakeman. at the 5 Bells, but it was Mr Holt that gave it me. (1702 Trials, Haagen Swendsen) Ball (1996: 246–247) states that the crucial change in the use of that took place in the second half of the 17th century: that began to be confined to nonpersonal antecedents and who to personal ones. This is, however, mainly in writing, while in spEModE, that is as frequent with personal as with nonpersonal antecedents, as shown by the figures presented in Table 49.3.
4.2 Who, whom and whose Who as a nominal relative meaning ‘whoever, anyone who’ is often used in Middle English (see Meier 1967: 281; Steinki 1932: 29; Poutsma 1926–1929: 985) but is not very frequent in the EModE period. Who as adnominal relative (i.e. in postmodifying relative clauses) first occurred with deities in the closing formulas of letters, with a personal name, and in nonrestrictive relative clauses (see Ryde´n 1983; Rissanen 1999: 294; Nevalainen and Raumolin-Brunberg 2002: 116–118). Not until the beginning of the 18th century, however, did who replace which with most types of personal antecedents (Ryde´n 1983: 132). Most examples of who in spEModE are with personal names, as in example (20). (20) Wil. I am not Beaux enough for that yet. To be short then, Beliria who you know lives with, and governs my Lady Young-Love, (1696 Drama, Manley, The Lost Lover) As subject with personal antecedents, who is considerably less frequent than that in my spEModE data (who: 24% and that: 69%, see Table 49.3 above). Who is common as subject with a personal antecedent only with proper names (see example 20 above) and in “progressive” (or “continuative”) relative clauses. Progressive relative clauses (see Ryde´n 1966: xlvii; Jespersen 1927: 83) are a type of nonrestrictive relative clause where the narrative is continued and expanded introducing a new element of action or a fact about the antecedent (see Ryde´n 1966: xlviii; Abbott 1966 [1870]: 176). In progressive clauses, who can be substituted by e.g. “and he”, as in examples (21)–(22).
783
784
V Early Modern English (21) RAWLINS ANSWERS. […] she said to me Madam I pity you, will no body Bail you? she told me I will send to my Brother who shall be Bail for you, (1702 Trials, Haagen Swendsen) (22) Mask. […] For your honest man,asI take it, is that nice, scrupolous, conscientious Person, who will cheat nobody but himself; such another Coxcomb, as your wise man, who is too hard for all the World, (1694 Drama, Congreve, The Double-Dealer) When who appears with a nonpersonal noun or an animal in spEModE (Table 49.3: four examples or 1%), some element of personification is involved. Personification of nonpersonal antecedents is found in literary EModE texts (e.g. in Shakespeare; cf. Abbott 1966 [1870]: 179–180; Franz 1924: 295). Example (23) is from Lyly’s comedy, where literary, or rather poetical, language is more frequent than in the other plays studied. (23) Alex. Not with Timoclea you meane, wherein you resemble the Lapwing, who crieth most where her neast is not. (1584 Drama, Lyly, Alexander and Campaspe) Whom and whose appear earlier than who as relativizers in Middle English and EModE. The use of the inflected forms could have favored the introduction of nominative/subjective who, to fill a gap in the paradigm (Steinki 1932: 30; Ryde´n 1983: 130). Whom, like who, is mainly used with reference to persons or human nouns but instances of whom with nonpersonal antecedents also occur, e.g. in Elyot’s writings from the early 16th century (Ryde´n 1966: 34–35). In my spEModE texts, whom occurs only with a personal noun (see Table 49.3) as in (24). (24) Capt. Harris, The next night I went to look for Sir George Barclay, whom the King told me I should certainly find by such a Sign of a White Handkerchief hanging out of his Pocket, (1696 Trials, Ambrose Rookwood) The possessive relativizer whose is mostly used with a personal antecedent in spEModE, as in example (25). (25) Host. To be scarce, is to be rare: and therefore where as he sayes Gentles whose wits be scarce, is as much as to say, Gentles whose wits be rare. (1599 Drama, Chapman, An Humerous Dayes Myrth) When nonpersonal whose occurs in spEModE, some implications of personal qualities or characteristics can be discerned (the nouns referred to include beauty and face). Historically, whose is the genitive of both personal (who) and nonpersonal (which) antecedents (see Campbell 1983: 292; Jespersen 1927: 116, 129).
4.3 Which Which could refer to both a personal and nonpersonal antecedent in EModE. With regard to written EModE and the early 16th century, Ryde´n (1966: 277) reports that in Elyot as many as 53% of the instances of which refer to a person. The works of Thomas Starkey (c.1535) and the Tyndale Bible (1525) show as much as 77% and 80% personal
49 Early Modern English: Relativization which, respectively. In the period 1600–1649, the figure for personal which is 10% (Dekeyser 1984: 71). The situation is different in the spEModE data used in this study: in only 5% of the examples does which refer to a personal antecedent (see Table 49.3: which as subject: 2% and as object 3%). My spEModE data does not reveal whether personal which decreases further. The few examples that occur are from the second period 1680–1719. Most of the examples of personal which are from Trials, as in example (26). (26) Mrs Busby. […] I saw a man in the Coach, which was Hartwel the Bayliff, (1702 Trials, Haagen Swendsen) Personal which occurs in witnesses’ speech as opposed to that of members of the legal profession. However, it is not possible to ascertain whether the use of which with persons in my spEModE data is typical of a particular social rank or speaker role (in Present-day English personal which occurs in nonstandard spoken language and in dialects, see Hughes et al. 2005: 29; Kjellmer 2002). Which tends to be much less frequent than that with all types of antecedent: nouns, names and pronouns. In spEModE, which seems to be frequent only where that is not used or for some reason is not the preferred relativizer. Which is used in sentential relative clauses with a clausal antecedent (where that does not occur, see Section 3.2) and in non-adjacent relative clauses, i.e. relative clauses that are at some distance from their antecedent. Which is then felt to be a more explicit construction than that as regards reference (Steinki 1932: 16–17; Rissanen 1984: 424; Jespersen 1927: 122). Furthermore, which is found with that + noun to avoid repetition of that, in e.g. that cloake which, that Matter which and in the combination that which (i.e. the nominal relative what in Present-day English). In adverbial relative clauses with a temporal or locative noun such as time, day and place as antecedent, which is the only alternative when a preposition precedes the relativizer. See example (27). (27) Lord Townshend. My Lord, having receiv’d Information that there was a Treasonable Correspondence carry’d on between the late Duke of Ormond, Duke D’Aumont, Coulange and Mr. Harvey, in which the Prisoner was concern’d, and was the Channel in which the Correspondence was convey’d; (1716 Trials, Francis Francia) The construction the which is extremely rare in my spEModE material. Only three examples occur and the relativizer is preceded by the preposition of. Two of the three examples are found in the Trial of the Duke of Norfolk, see example (28). Individual writers between 1520 and 1560 prefer the which but it is not frequent generally (see Ryde´n 1966: 278–279). (28) Attorney. […] so that there was for every Name a Cypher, of the which 40 was for the Duke, and 30 for the Lord Lumley. (1571 Trials, Thomas Howard Duke of Norfolk) Of which as an alternative to whose with nonpersonal antecedents is first recorded in Middle English (see Johansson 2002; Schneider 1993). Of which does not replace its main rival whereof until the second half of the 17th century and it is not a frequent alternative to whose in the EModE period (Schneider 1993: 253). Only one example of of
785
786
V Early Modern English which occurs in my spEModE texts, see example (29). However, the emergence and use of the of which construction is important as an example of the desire in EModE times to mark more clearly the personal/nonpersonal contrast. Whose is increasingly regarded as the possessive relativizer only for a personal noun (cf. Dekeyser 1984: 70–71). (29) Lord Townshend. […] was after that Letter had been deliver’d to me, the Contents of which will sufficiently justifie the Precautions used in that Warrant. (1716 Trials, Francis Francia)
4.4 The zero construction This construction occurs with both personal and nonpersonal antecedents in spEModE but is most frequent with the nonpersonal category. As is the case with that, the zero relativizer is particularly frequent with certain types of antecedent: nothing (as in example 30), something, any thing, all, the general noun thing and with a superlative + noun. (30) Bell. I have nothing Ø I can call my own; you, like a bold Invader, have born away by force what else my Faith and Love had offer’d. (1719 Drama, Killigrew, Chit-Chat) Furthermore, a zero relativizer is often used in prepositional constructions with a stranded preposition (see Section 5.2), in cleft sentences, as in example (31), and in adverbial relative clauses, mainly with time expressions, as in example (32). (31) Flor. But since such a one is not upon Earth, I shall insist ’tis in Resistance Ø we acquire our Fame. (1719 Drama, Killigrew, Chit-Chat) (32) Town. […] The Day Ø we din’d last together, Bellamar negligently went into the Room before Lurcher, sate upon his Right Hand at Table. (1719 Drama, Killigrew, Chit-Chat) In my spEModE material, the zero construction is used more frequently in the Drama texts than in the Trials. It seems as if the zero construction is one feature that individual playwrights want to include in their characters’ dialogue. As mentioned earlier in this paper, there is an important difference between the two spEModE periods here: the zero relativizer is nearly twice as frequent in the period 1680–1719 (see Table 49.1 above). It is of course not possible to maintain that the increase in the use of the zero construction mirrors real dialogue at the time but the playwright may just be anxious to render EModE speech as closely as possible. Drama is more speech-related in some respects, as interruptions and lexical repetitions reveal (see Culpeper and Kyto¨ 1999: 293–312, 2000: 179–185). In Killigrew’s comedy Chit-Chat (1719), the zero construction is used very often by characters representing the higher social classes. In this comedy, the zero relativizer is used with most of the typical antecedents of zero (e.g. person(s), people, thing, nothing) and in clefts and adverbial relative clauses, see examples (31)–(32) above. Individual writers and letter writers use zero, but it is rare in the early 16th century (Ryde´n 1966: 286, 294–295, cf. also Table 49.1 above). When the zero relativizer functions as object (direct + prepositional) with nonpersonal antecedents in spEModE, it is slightly more frequent (36%) than both that (31%)
49 Early Modern English: Relativization and which (33%, see Table 49.3). When, on the other hand, the zero construction functions as subject, it is rare and mainly restricted to certain syntactic environments. This will be discussed in the next section, which deals with how the subject function of the relativizers is realized (see Section 5.1).
5 Forms and functions of the relativizers This section will focus on the forms and functions of EModE relativizers, more precisely, whom and zero as subject, and preposition + relativizer as prepositional object.
5.1 Whom and the zero construction as subjects In examples (33) and (34) below, whom is the subject of the main clause and at the same time the object of the inserted clauses he then well knew and they said. Such relative clauses are a type of blended construction (or “push-down relative clauses”; see Quirk et al. 1985: 1118–1120; Quirk 1957: 103 for Present-day English examples). As is the case in Present-day English, there was in EModE, a tendency to regard whom as the direct object of the verb in the main clause (see Visser 1963: 495–496). Jespersen (1927: 198–199) refers to “the speech-instinct”, which does not readily accept two subjective forms in the same clause. As shown in Table 49.3, there are only four examples in my spEModE texts. Hence it is hardly possible to argue that subjective whom is a frequent feature of the spoken language. (33) Serj. […] his Practice to join himself in marriage with the Scotish Queen, whom he then well knew falsely to claim and pretend Title to the present possession to the Crown of England: (1571 Trials, Thomas Howard Duke of Norfolk) (34) Mrs Rawlins. There was a Minister in the House, whom they said had been there about a quarter of an hour but I supposed longer, (1702 Trials, Haagen Swendsen) As with subjective whom, examples of the zero construction as subject are not very frequent in my spEModE data (Table 49.3: 3% for both personal and nonpersonal antecedents). Zero occurs mainly in presentative constructions, such as ’tis, who is it … (in cleft sentences), there is and here is (Ryde´n 1966: 268–269; Curme 1931 235–236; Poutsma 1926–29: 995, 1001). See examples (35) and (36). Zero as subject could be taken as a feature that is more typical of EModE speech than of literary language. The zero relativizer as subject occurs for instance in Shakespeare (cf. Abbott 1966 [1870]: 164; Barber 1997: 115–116), but is extremely rare in Elyot’s works and those of his contemporaries in the early 16th century (Ryde´n 1966: 278, 290). (35) L. C.J.H. Who was it Ø came to you? (1702 Trials, Haagen Swendsen) (36) Scrub. […] There’s not Day Ø goes over his Head without Dinner or Supper in this House. (1707 Drama, Farquhar, The Beaux Stratagem) Zero as subject does not, however, increase in the second spEModE period (1680–1719), as is the case with the zero construction as object and prepositional object.
787
788
V Early Modern English
5.2 Preposition + relativizer as prepositional object As in Present-day English, prepositional objects (and adverbials) in EModE could be expressed in different ways, depending on the placement of the preposition; both preposition + whom/which (“pied piping”) and whom/which/that/zero + preposition (“stranding”) occur. First, it should be noted that the object function is one of the few functions where that is not the most frequent form in spEModE. Table 49.3 above indicates that with personal antecedents, whom is most frequent (41%), whereas the most common relativizer with a nonpersonal antecedent is the zero construction (36%). If only prepositional objects are considered, that (+ a stranded preposition) is, again, the most frequent form with both types of antecedent, 50% (see Table 49.4). When a wh-form is used, pied piping is the most common pattern. With whom it is the only pattern, whereas which occurs with a stranded preposition in a few examples (see Table 49.4). What is most interesting, however, is that stranding is nearly twice as common (63%) as preposition + relativizer (37%) in my spEModE material. Table 49.4: Prepositional objects in spEModE: placement of preposition Function
Prepositional object Pied piping
Stranding
Total
That Whom Which Zero
— 8 (26%) 23 (74%) —
26 (50%) — 8 (15%) 18 (35%)
26 8 31 18
Total
31 (37%)
52 (63%)
83 (100%)
A construction with a stranded preposition was not generally looked upon as informal style in the EModE period. It is at the beginning of the 18th century that some writers started questioning the correctness of preposition stranding (see Visser 1963: 402–403). The fact remains, however, that stranding is the preferred pattern in spEModE and possibly it was or became associated more with spoken than with written EModE. In example (37), stranding with Ø and that is illustrated. (37) Mr. Conyers. My Lord, the first Meeting Ø Mr. Porter speaks of, where the Prisoner Mr. Rookwood was, is at the Globe-Tavern, where this Discourse was; the next Meeting that he speaks of, was on the Friday Night, before the first Saturday […] (1696 Trials, Ambrose Rookwood) Preposition + which is the predominant pied piping pattern. It occurs in 74% of the pied piping constructions (see Table 49.4). Often preposition + which seems to be obligatory due to the complexity of the sentence structure, as in example (38). Ryde´n (1966: 297) states that preposition + which is preferred by individual writers in the early 16th century (e.g. Sir Thomas More and William Tyndale). (38) Do. […] that caulfe with a white face is his faire daughter, with which, when your fields are richly filled, then will my race content you, (1599 Drama, Chapman, An Humerous Dayes Mirth)
49 Early Modern English: Relativization
6 Summary This paper describes the use of relativizers in EModE, focusing on speech-related Trials and Drama texts (spEModE). The periods studied are 1560–1599 and 1680–1719. What is most striking as regards the use of relativizers in my spEModE material is the predominance of the relativizer that. Although the wh-forms start competing with that also in restrictive relative clauses (mainly in writing), that does not decrease in frequency to any great extent. That is used with all types of personal and nonpersonal antecedents, nouns, proper names, and pronouns. It occurs in both restrictive and nonrestrictive relative clauses but decreases in nonrestrictive relative clauses in the second spEModE period (1680–1719). Who does not really threaten the position of that as subject with personal antecedents in spEModE: it is used in only 24% of the examples (often with proper names). Instead, it is which that disappears more and more with a personal antecedent. The real threat to the position of that in spEModE is the zero construction, which increases and is nearly twice as frequent in the second spEModE period (1680–1719). The zero construction is more common with nonpersonal antecedents than that but is used as frequently as that with antecedents such as all, nothing, and a superlative + noun. The function of direct object with nonpersonal antecedents is the only function where that is not predominant in my spEModE texts. Here, the zero construction is the most frequent relativizer. At the end of the EModE period studied in this paper, we recognize many of the features of Present-day English usage of relativizers. These features probably appeared earlier in spEModE texts: which is more or less restricted to nonpersonal antecedents, that occurs mainly in restrictive relative clauses where it is particularly common with all, nothing, thing and superlatives + noun. Also, the zero construction increases in frequency. However, it is important to note that in spEModE around 1700, who is not yet a frequent alternative to that with personal antecedents. Acknowledgments: First of all, I want to thank Mats Ryde´n for valuable comments on my paper and for taking a special interest in my work. I also want to thank Angela Falk, Christer Geisler, Edward Long, and Terry Walker for valuable comments, and I am grateful to Christer Geisler and Merja Kyto¨ for their help in suggesting and providing the material from A Corpus of English Dialogues.
7 References Abbott, E. A. 1966 [1870]. A Shakespearian Grammar: An Attempt to Illustrate Some of the Differences between Elizabethan and Modern English. New York: Dover Publications. Ball, Catherine N. 1996. A diachronic study of relative markers in spoken and written English. Language Variation and Change 8: 227–258. Barber, Charles. 1997. Early Modern English. 2nd edn. Edinburgh: Edinburgh University Press. Biber, Douglas. 1988. Variation across Speech and Writing. Cambridge: Cambridge University Press. Campbell, Alistair. 1983. Old English Grammar. Oxford: Oxford University Press. Culpeper, Jonathan and Merja Kyto¨. 1999. Modifying pragmatic force: Hedges in Early Modern English dialogues. In: Andreas Jucker, Gerd Fritz, and Franz Lebsanft (eds.), Historical Dialogue Analysis, 293–312. Amsterdam/Philadelphia: John Benjamins. Culpeper, Jonathan and Merja Kyto¨. 2000. Data in historical pragmatics: Spoken interaction (re)cast as writing. Journal of Historical Pragmatics 1(2): 175–199.
789
790
V Early Modern English Culpeper, Jonathan. 2001. Language and Characterisation. People in Plays and Other Texts. London: Pearson Education/Longman. Curme, George O. 1931. Syntax. Vol. III: A Grammar of the English Language. Boston: Heath. Dekeyser, Xavier. 1984. Relativizers in Early Modern English. A dynamic quantitative study. In: Jacek Fisiak (ed.), Historical Syntax, 61–87. Berlin/New York: Mouton de Gruyter. Franz, Wilhelm. 1924. Shakespeare-Grammatik. Heidelberg: Winter. Hughes, Arthur, Peter Trudgill, and Dominic Watt. 2005. English Accents and Dialects. 4th edn. London: Hodder Arnold. Jacobsson, Bengt. 1994. Nonrestrictive that-clauses revisited. Studia Neophilogica 66(2): 181–195. Jespersen, Otto. 1927. A Modern English Grammar on Historical Principles. Part III. Syntax (Second Volume). Copenhagen: Munksgaard. Johansson, Christine. 2002. The relativizers whose and of which in Middle English. In: Poussa (ed.), 37–49. Johansson, Christine. 2006. Relativizers in nineteenth-century English. In: Merja Kyto¨, Mats Ryde´n, and Erik Smitterberg (eds.), Nineteenth-century English: Stability and Change, 136–182. Cambridge: Cambridge University Press. Kjellmer, Go¨ran. 2002. On relative which with personal reference. Studia Anglica Posnaniensia 37: 20–38. Kyto¨, Merja and Jonathan Culpeper. 2006. A Corpus of English Dialogues 1560–1760. With the assistance of Terry Walker and Dawn Archer. Uppsala University and Lancaster University. http://www.engelska.uu.se/Research/English_Language/Research_Areas/Electronic_Research_ Projects/A_Corpus_of_English_Dialogues Kyto¨, Merja and Terry Walker. 2006. Guide to A Corpus of English Dialogues 1560–1760. Uppsala: Acta Universitatis Upsaliensis. Meier, H. M. 1967. The lag of relative who in the nominative. Neophilologus 51(3): 277–288. Nevalainen, Terttu and Helena Raumolin-Brunberg. 2002. The rise of relative who in Early Modern English. In: Poussa (ed.), 109–121. Poussa, Patricia (ed.). 2002. Relativization on the North Sea Littoral. Mu¨nchen: Lincom Europa. Poutsma, Hendrik. 1926–1929. A Grammar of Late Modern English. Groningen: Noordhof. Quirk, Randolph. 1957. Relative clauses in educated spoken English. English Studies 38: 97–109. Quirk, Randolph, Sidney Greenbaum, Geoffrey Leech, and Jan Svartvik. 1985. A Comprehensive Grammar of the English Language. London: Longman. Rissanen, Matti. 1984. The choice of relative pronouns in 17th century American English. In: Fisiak (ed.), 417–435. Rissanen, Matti. 1999. Syntax. In: Roger Lass (ed.), The Cambridge History of the English Language. Vol. III. 1476–1776, 292–301. Cambridge: Cambridge University Press. Ryde´n, Mats. 1966. Relative Constructions in Early Sixteenth Century English. With Special Reference to Sir Thomas Elyot. Uppsala: Acta Universitatis Upsaliensis. Ryde´n, Mats. 1970. Coordination of Relative Clauses in Sixteenth Century English. Uppsala: Acta Universitatis Upsaliensis. Ryde´n, Mats. 1974. On notional relations in the relative clause complex. English Studies 55: 542–545. Ryde´n, Mats. 1983. The emergence of who as relativizer. Studia Linguistica 37: 126–130. Ryde´n, Mats. 1984. Na¨r a¨r en relativsats no¨dva¨ndig? [When is a relative clause essential?]. Moderna Spra˚k 78: 19–22. Schneider, Edgar. 1993. The grammaticalization of possessive of which in Middle English and Early Modern English. Folia Lingustica Historica 14(1.2): 239–257. Steinki, Johannes. 1932. Die Entwicklung der Englischen Relativpronomina in Spa¨tmittelenglischer und Fru¨hneuenglischer Zeit. Breslau: Die Schleschische Friedriech-Wilhelms-Universita¨t: Dissertation. Visser, F. Th. 1963. An Historical Syntax of the English Language. Vol. I. Syntacical Units with One Verb. Leiden: E. J. Brill.
Christine Johansson, Uppsala (Sweden)
50 Early Modern English: Literary language
791
50 Early Modern English: Literary language 1 2 3 4 5
Introduction The construction of literary language Literary language in its communicative context Summary References
Abstract Literary language between 1500 and 1700 was strongly influenced by classical models of poetic language from which it imported figures, tropes, and schemes to create a high or grand style for literary genres. Although not all literary language was written in high style, the high style shaped patterns of syntax and lexical usage in literary texts by encouraging copious, excessive, and elaborately ornamental usage. The Early Modern impulse to codify and organize English was reflected in a self-conscious approach to literary language: the period saw the publication of many rhetorics designed to develop and classify literary usage. Literary texts also contributed to English nation-building efforts during the 16th and 17th centuries; efforts to develop a language of literature sought to legitimize English as a medium for artistic expression.
1 Introduction To define literary language in the Early Modern period – indeed, in any period – is to engage in a circular practice: on its most basic level, literary language is language characteristic of literary genres, but literary genres are identifiable as those genres that employ literary language. Moreover, as the Russian formalist Yury Tynyanov (2000 [1924]: 29–49) pointed out, works themselves only resonate within genre systems, and to take a literary work out of its generic context is to change its literary function. What makes literary language literary, then, is our knowledge of its generic function and the place of the genre in a culturally-negotiated hierarchy of aesthetic value. (For further description of thorny problems in defining literary language, see Herman 1983 and Carter and Nash 1983.) In order to describe language that is distinctly literary, we find our focus falling inevitably upon usage that is marked in some way – in other words, to describe the features that particularly constitute linguistic art, we attempt to isolate the ways that it differs from other registers. Sixteenth-century rhetorician George Puttenham, for example, describes the ways that literary figures (the tools for poetic ornamentation) depart from ordinary speech in his Arte of English Poesie (1589), “As figures be the instruments of ornament in every language, so be they also in a sort abuses [distorted lexis], or rather trespasses [distorted syntax], in speech, because they passe the ordinary limits of common utterance” (Puttenham 2007 [1589]: 238). Puttenham identifies figures as distortions of ordinary syntax and lexicon, and it is, in a sense, this linguistic distortion that we most often see when we examine literary language. Although it is an oversimplification to strictly reduce our understanding of literary language to a set of marked features, any description of literary language necessarily foregrounds some aspects or variables as “literary.” “Literary language”, in Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 791–807
792
V Early Modern English other words, is a perceptual category, and to study it is to study our perceptions of literariness. For this reason, any examination of Early Modern literary language poses some historical dilemmas of taxonomy. The very phrase literary language is anachronistic when applied to the period between 1500 and 1700: the terms literary and literature do not gain their modern associations with writing in a creative, imaginative, or artistic fashion until the 18th and 19th centuries. Users of Early Modern English certainly had ideas about artistic uses of language, of course, some of which we no longer hold. When 16th-century rhetorician Alexander Gil, for instance, names the six major dialects of English as the general, the Northern, the Southern, the Eastern, the Western, and the Poetic, he nicely calls to our attention the differences between Early Modern ideas of literary language and our own (Gil 1621: 6). Apparently Gil considered poetic language to be identifiable by notable features in the same way that one might characterize a regional dialect. Few poets today, though, would consider their language to be distinct from “general” or regional English, let alone part of a separate (but communally-held) dialect. An obvious question then arises when considering 17th-century poets and theorists such as Gil: if literary language is a perceptual category (as per above), whose perceptions should we examine, theirs or ours? In this section, I attempt to negotiate between these two endeavors, examining a wider variety of uses of literary language than Puttenham and Gil perhaps had in mind, but acknowledging wherever possible Early Modern ideas about language as artistic craft. Literary language in the early Modern period carries over many of the practices of literary composition from the late Middle Ages, certainly, but it becomes further conventionalized and further developed in the Renaissance, and it is more widespread, more self-conscious, and, as we will see, tied to the growth of English nationalism. The rise of literary language also developed through a confluence of material factors in the 16th century: a radical increase in literacy brought about by the multiplication of grammar schools, the decreasing use of French and Latin in formal settings, and the marked increase in the availability of books owing to cheap paper and the establishment of the printing industry in London. For comparison, we find 54 titles in 1500, but 214 by 1550, and 577 by 1600 (Go¨rlach 1991: 5–7). The Early Modern period saw the systematizing and theorizing of the words and style of literary English, and the literary register played a larger sociopragmatic role in cultural shifts in England, such as the spread of literacy and the reformulation of English literature as a nationalist project. To examine the language of literature, therefore, this chapter will look at what characterizes the construction of literary language (its syntax and lexicon), and will then discuss the importance of literary language in a larger communicative context (the function of the developing sense of national literature, the pragmatics of language play).
2 The construction of literary language Our perspectives on literary language are often discussed in terms of registers, or levels of style. What has been called “the high style” or “the grand style” has received particular attention as being the culmination of ornate and ornamental stylings, and was the one most clearly recognized during the Early Modern period as being elevated and suitable for literary purposes (Adamson 2001). Thomas Wilson, in his Arte of Rhetorique,
50 Early Modern English: Literary language called it, “the great or mightie kinde [style], when we vse great wordes, or vehement figures” (Wilson 1909 [1560]: 169). Middle and low styles are present and important for literary writing as well, certainly, in their own right and as foils for high style. In fact, our attention to high style has been disproportionate enough to influence the way that we approach authors and works of the period. A. J. Gilbert (1979: 173) discusses the way that our emphasis on John Donne’s high style, for example, has resulted in suppression of low style in editions of his works and a resulting impression of his prose as more elaborate than it actually is. Our emphasis on the high style has also affected the genres that we focus upon: the heroic style of epic and the flowery language of love poetry have been emphasized at the expense of the lyrics of pastoral poetry which are typically written in lower style. (For a contemporaneous, i.e. late 16th century, discussion of types of poetry, see Philip Sidney’s Apology for Poetry [1970 (1595)].) Yet the spotlight of literary language study falls most prominently on high style because it is the most distinct and most artificial (in the sense of constructed artifice) form of literary language and thus most apparent to observers both in the period and afterwards. The focus on high style (both by Early Modern critics and present-day ones) occurs because it is the most marked of literary forms (as discussed above), and my discussion here will similarly concentrate on this style as being particular to and characteristic of literary language, even though literary language certainly did not use it exclusively. High style is built around classical models, unusual constructions, involved syntax and noticeably decorative rather than mimetic style. Consider the famous sixteen-line first sentence of Paradise Lost, (1): (1)
OF Mans First Disobedience, and the Fruit Of that Forbidden Tree, whose mortal tast Brought Death into the World, and all our woe, With loss of Eden, till one greater Man Restore us, and regain the blissful Seat, Sing Heav’nly Muse, that on the secret top Of Oreb, or of Sinai, didst inspire That Shepherd, who first taught the chosen Seed, In the Beginning how the Heav’ns and Earth Rose out of Chaos: Or if Sion Hill Delight thee more, and Siloa’s Brook that flow’d Fast by the Oracle of God; I thence Invoke thy aid to my adventrous Song, That with no middle flight intends to soar Above th’ Aonian Mount, while it pursues Things unattempted yet in Prose or Rhime (1667 Milton, Paradise Lost ll. 1–16)
The inverted structure of this sentence begins with an extended prepositional phrase, and progresses through a set of extended and embedded relative clauses, yoking syntactic features for an effect that is quite distinct from spoken English. The sentence serves as a distillation of the features of high style: classical models (invocation of the muse), involved syntax, use of rhetorical figures (apostrophe), use of abbreviations (Heav’nly, advent’rous), conservative morphosyntax (thee, thy). The passage also declares as well as performs its desire to syntactically rise above the common language – no “middle
793
794
V Early Modern English flight” will do when it “pursues / Things unattempted yet in Prose or Rhime”. The literary projects of the Early Modern period developed and refined this ambitious, highflying style that Milton describes, and although it was certainly not the only form of literary language, it was one of the most noticeable and consciously constructed. The very term high style points to a hierarchy that privileges particular types of language: written language, formal language, the language of the upper classes, the language of the educated. We see contemporaneous attitudes when we look at the coexistence of high and low styles in drama, in which the upper class characters often speak in blank verse and high style oratory, and the lower class characters speak in low style and in prose. Our attitudes towards high style (both the term for it and also the register itself) have shifted over time and altered the goals and construction of literary language. In the last two centuries, our ideals of literary language have grown out of regular speech (cf. Wordsworth’s “language really used by men”), and our most literary of works (increasingly novels) are fed by our notions of everyday language. Present-day readers hold it up as a great virtue of literary writing that characters should speak like real people (though, of course, our ideas of what real people sound like remain cultural artifice). In the Renaissance, by contrast, the features of high style followed classical models and were determined by their distinctness from the everyday. This affected both writers who were influenced by ideals of high style and writers (especially later in the 17th century) who rejected them and defined their own styles in opposition.
2.1 Classical models Renaissance critical practice – and thus, its models for literary composition – drew from classical templates and was primarily directed towards tropes and figures of thought and writing. As the central categories for understanding literary and rhetorical methods of discursive organization, these figures, schemes, and tropes included a broader range of stylistic features than our present-day lists of “figures of speech” take into account. It is difficult for present-day readers to inhabit the mindset required for the 16th century rhetorically inflected literary experience, and to understand the practice of the reader who read Sidney’s Arcadia with pen in hand, labeling all of the figures and tropes in the margin. (See the frontispiece to Adamson et al.’s 2007 volume for an image of a page of Arcadia marked in this manner.) And while not every reader had equal training, critical reading entailed conversance with dozens of figures (Henry Peacham 1577, for instance, lists nearly 200). Rhetorics of the 16th and 17th century were essentially lists categorizing figures, schemes, and tropes and illustrating their use (even if the relationships between the three were differently negotiated in different sources). The classical categories were imported by Early Modern rhetoricians and adapted in different ways. Bernard Lamy (1676), for example, differentiated between figures and tropes based on functionality. He describes tropes as “an Ornament to Discourse” (Lamy 1676: 90) and figures as attempts to express passion in language; they are “Manners of Speaking, different and remote from the ways that are ordinary and natural; that is to say, quite other than what we use, when we speak without passion” (Lamy 1676: 94–95). Abraham Fraunce (1588: 3) constructs a provenance for tropes in The Arcadian Rhetorike, asserting the tropes were invented because of lexical inadequacies, but continued “by reason of the delight and pleasant grace thereof ”.
50 Early Modern English: Literary language Figures are the largest category, and sometimes the superordinate term for all of the techniques, but there can be no exhaustive list of Renaissance figures because rhetoricians were not in complete agreement on what would belong on such a list. I will not attempt to delineate and illustrate the use of figures here (for the clearest categorizing description see the works of Sylvia Adamson, especially Adamson 1999). But the long lists of Greek and Latin figures (from anadiplosis to zeugma) coalesce around two general categories: figures of varying and figures of amplifying, and these two divisions indicate the richness and amplitude of Early Modern literary style. Figures of varying include ways to multiply words; rhetoricians itemized examples of paronomasia (words that begin with the same sound, but have contrasting meanings), polyptoton (repetition of the same root word with varying affixes), paraphrasis (circumlocutory and excessive discourse), and diaeresis (the use of particulars to stand in for general categories). Figures of amplification were designed for elaboration and exaggeration. John Hoskyns described their function in Directions for Speech and Style: “To amplify and illustrate are two the chiefest ornaments of eloquence, and gain of men’s minds two the chiefest advantages, admiration and belief ” (Hoskyns 1935 [1599]: 17). Although figures functioned as the dominant analytic tool for understanding literary style and rhetorical construction, they eventually began to be viewed with suspicion (as we saw in the quotation from Puttenham in Section 1). Figures and tropes, it was argued, distort ordinary language, they bend the truth, they present reality in an altered way in service of persuading and pleasing the reader, and as such they began to be regarded as potentially morally questionable. Reformers of usage in the 17th century in the cause of science called for the rejection of tropes as tools of falsification: Bacon spoke of them in The Advancement of Learning (1605) as “deceiving expectation” (II.V.3), and Thomas Sprat (1667) in his History of the Royal-Society (1667) wrote, “Who can behold without indignation how many mists and uncertainties these specious Tropes and Figures have brought on our Knowledg?” (Section 20; both cited in Partridge 1971: 20). Bernard Lamy warns of the dangers that figures, as expressions of our passions, can also work public feeling in ill ways: “They are Instruments used to shake and agitate the Minds of those to whom we speak: If these Instruments be managed by an unjust Passion, Figures in that Man’s Mouth are like a Sword in the Hand of a Mad Man” (Lamy 1676: 142). Figures and tropes, therefore, were the tools of literary language, but they created anxiety owing to the perceived possibility of abuse. If literary language has the power to inspire us, to move us, to please us, to touch our hearts, it also has the power to incite us, to enrage us, to indulge our destructive emotions. In the wrong hands, literary language seems a dangerous thing. Striking a balance between levels of style also became more fraught as the period wore on. Navigating between extravagant and barren style has always been a point of cultural taste, and the mounting 17th-century condemnations of over-the-top style merely present a shift of the general taste, and not a complete rejection of all of the elements of Renaissance high style. Dudley North’s (1645) A Forest of Varieties presents a typical example. He states, “The Poetry of these times abounds in wit, high conceit, figure, and proportions; thinne, light, and emptie in matter and substance; like fine colored ayery bubbles or Quelque-choses, much ostentation and little food” (North 1645: 2), but a few pages later he asserts that we must be sure to elaborate enough because otherwise we risk obscurity: “I feare wee all often unwillingly incurre the errour of it by thinking our meaning as open to others, as to our selves,
795
796
V Early Modern English when indeed the Characters of our expression are fully supplyed by our owne understanding to our selves, whilst to others they are lamely contracted and imperfect” (North 1645: 4). North asserts that we should avoid empty overwriting and sparse underwriting, but he doesn’t provide a list of criteria for how we are to distinguish between the two. Ben Jonson also describes the two extremes of style in his early 17th century precepts on style; his description of the body of a language includes a description of the flesh, blood, and bones of style. Fleshy style, he describes, has more than enough and is “fat and corpulent” through, for example, periphrasis. The opposite is also a problem: “but where that wanteth, the Language is thinne, flagging, poore, starv’d; scarce covering the bone, and shewes like stones in a sack. Some men, to avoid Redundancy, runne into that; and while they strive to have no ill blood, or Juyce, they loose their good. There be some styles againe, that have not lesse blood, but lesse flesh, and corpulence” (Jonson 1640: 121). Striking the middle path between fleshy style and style that is “thin, flagging, poor, starved, scarce covering the bone”, of course, is a point of taste, and different authors have achieved it in different ways. Modern readers find the style of 16th and 17th century literary texts to be more complex and ornate than our present tastes dictate. But many of the issues and choices of style remain the same.
2.2 Syntax and lexicon The features of high style shaped the syntax of Early Modern literary language, pulling the sentential organization in the direction of involved syntax and encouraging particular constructions. We see in literary language, for example, an extensive use of subordination, parallelism, balanced clauses, modification, and coordination. These manifest differently in verse and in prose, of course, but the use of modifying words and clauses resulted in some of the long descriptive sentences that we find in Early Modern prose, as in (2). (2)
Turning from the table, she discerned in the roome a bed of boughes, and on it a man lying, depriued of outward sense, as she thought, and of life, as she at first did feare, which strake her into a great amazement: yet hauing a braue spirit, though shadowed vnder a meane habit, she stept vnto him, whom she found not dead, but laid vpon his back, his head a little to her wards, his armes foulded on his brest, haire long, and beard disordered, manifesting all care; but care it selfe had left him: curiousnesse thus farre affoorded him, as to bee perfectly discerned the most exact peece of miserie; Apparrell hee had sutable to the habitation, which was a long gray robe (1621 Wroth, The Countesse of Mountgomeries Urania 3)
We see, here, both coordinate and subordinate clauses, pre- and post-modifications (great amazement, braue spirit, haire long, beard disordered), and inverted syntax (Apparrell hee had). Lady Mary Wroth’s The Countess of Montgomery’s Urania is perhaps not a representative literary work in many ways, but a sentence like (2) shows how the full stop (period) organized literary prose less syntactically and more rhetorically. The punctuation indicated pauses of varying lengths – ordering the oral delivery rather than the logical structure, though the functions certainly overlap). Such sentences can
50 Early Modern English: Literary language therefore seem uncomfortably long to present-day readers. The desire for amplification also resulted in, for example, compounding, as in (3): (3)
a. With joy and jollitie needs round must rove (1642 More, Psychodia platonica 96) b. Viewing Leanders face, fell downe and fainted (1598 Marlowe, Hero and Leander sig. C4v.)
Example (3) shows syntactic doublings with the parallels emphasized through the alliteration. These were popular and sometimes formulaic (The phrase “joy and jollity” in (3a), for example, retrieves 23 hits in Early English Books Online [Chadwyck-Healey 2003–11]). Decorative writing also used words in unusual ways for aesthetic effect. Early Modern English in general was characterized by an increase in functional shift – the conversion of a word from one functional category to another (i.e. noun to verb) – and literary writing particularly exploits the flexibility in morphosyntactic categories. The functional shift derived from the simultaneous liberation of structure from the system of inflectional case endings and the shifting of word order patterns such that synthetic constructions were declining in favor of analytic ones (Rissanen 1999: 187). This looser structure unleashed lexical items from their strict functional categories, and writers were freer to use them in other functional ways: (4)
a. Cruell and sodaine, hast thou since Purpled thy naile, in blood of innocence? (1633 Donne, Poems 119) b. Thy saints trust in thy name, Therin they joy them: (1823 [c.1599] Mary [Sidney] Herbert, Countess of Pembroke, Psalm 52: 93)
In the first instance, Donne uses purple as a verb, and in the second Mary (Sidney) Herbert uses joy as a verb. These creative functional shifts were one of the hallmarks of Renaissance literary style. Stylistic expansion happens through doublings, through functional shift, and through elaborations. Some elaborating stylistic strategies came into vogue in the 1580s in the wake of John Lyly’s (1578) Euphues. The influence on style was so marked that the term Euphuistic arose for a style based on comparisons (sometimes reinforced by phonetic cues like alliteration). Here is Lyly’s description of his hero: (5)
This younge gallant, of more wit than wealth, and yet of more wealth than wisdome, seeing himselfe inferiour to none in pleasant conceits, thought himself superiour to al in honest conditions … (1578 Lyly, Euphues 1)
The syntactic parallelism and semantic patterning – the comparisons (more wit than wealth), the relational words (inferiour to none, superiour to al ) – were a runaway sensation in the 1580s. Notice here, also, the use of figures: antithesis (inferiour to none) and isocolon (more wit than wealth, more wealth than wisdom). Euphuistic style had a short run, however, and quickly turned from a popular phenomenon into the object of parody. The very possibility for fads in literary style, though, shows the
797
798
V Early Modern English emerging significance of English literary language, and the readiness of the community of readers for experimentation and wide-ranging transformation in English literary style. Examining the words of literature is also critical to understanding literary language, and important for our understanding of the Early Modern period in general. Literary texts have been one of the primary sources for our knowledge of the changing lexicon of English in the 16th and 17th centuries: they have been disproportionately represented in the OED, and they constitute a disproportionate slice of extant evidence. And yet we know that literary language is not colloquial language. So examining the words of literature reveals both contemporaneous ideas about the ways that literature may be used as a source for lexical change and about how word choice is related to high style. James Beattie defines the “English poetic dialect” in 1762 as consisting of seven characteristics: (1) Greek and Latin idioms (quenched of hope), (2) words with extra syllables (dispart), (3) ancient words ( fealty), (4) uncommon words (cates), (5) abbreviated words (o’er), (6) compound epithets (rosy-finger’d), (7) nouns transformed into verbs and participles (cavern’d) (cited in Sherbo 1975: 1). This analysis, though written after the period in question, was composed largely with Renaissance poetry in mind, so it gives us a fairly good idea of the features of literary (especially poetic) language that stood out for its users and readers. The list itemizes those we might expect: the out-of-the ordinary words, the fancy or elaborate ones: borrowed words, old words, dressed-up words. These are the lexical items that are particularly marked in literary usage and they correspond to (or serve as) stylistic figures. One prominent aspect of the Renaissance literary lexicon (and an important influence upon word choice) lies in the emphasis on copiousness, an aspect of high style. Erasmus’s De Copia (1963 [1512]) led the way in describing the abundant or copious style – providing as an example of its methods 144 different ways to say “Thank you for your letter”. This verbal excess went along with the figure of synonymia and notions of abundance. Consider Peacham’s definition and example: “Synonimia, when by a variation and change of words, that be of lyke sygnifycation, we iterate one thing divers tymes […] [s]ometime with words, thus. / Alas many woes, cares, sorrowes, troubles, calamities, vexations, and miseries doe beseige me round about […]” (Peacham 1577: sig. P4r). The driving philosophy seems to be that if one word would do, then two words must be better. Adamson describes the 16th-century prominence of synonymia and dates its decline to 1600. After this point, it begins to be considered one of the “vices of style” (Adamson 2007: 18). But through the 16th century, a superfluity of vocabulary was required for literary genres – a stylistic preference which pushed the language not to merely expand its technical and specialized vocabulary, but to expand its general lexicon as well. A poetic lexicon was also characterized by its excess in degree: beautiful language is ornate and embellished, and so word choice is exaggerated. The use of figures of amplification previously discussed influenced literary lexical choice. Hyperbole, for example, is a figure that encourages lexical diversity and lexical superfluity, as Hoskyns illustrates with Sidney’s Arcadia: (6)
The world should sooner want occasions than he valour to go through them (29) (cited in Hoskyns 1999 [1599]: 413)
50 Early Modern English: Literary language Other amplification figures center on word opposition, or the choice of words in context, such as synoeciosis, a type of contradiction in terms: (7)
A wanton modesty, and an enticing soberness (315) (cited in Hoskyns 1999 [1599]: 415)
These excessive words were characteristic of literary language, and certain genres in particular. A passage from Edward III that is widely attributed to Shakespeare mocks over-the-top words in love poetry: (8)
Better then bewtifull thou must begin, Deuise for faire a fairer word then faire, And euery ornament that thou wouldest praise, Fly it a pitch above the soare of praise, For flattery feare thou not to be conuicted, For were thy admiration ten tymes more, Ten tymes ten thousand more [the] worth exceeds Of that thou art to praise, [thy] praise’s worth, (1596 Shakespeare(?), Edward III 13)
In this passage, Edward III wants words that are bigger, better, and more beautiful than the typical words for describing beauty. The speaker asserts that one could never overstate the beauty of the referent, a parody of the vocabulary of amplification. And the passage exemplifies the poetic ideals that it describes, in the use of polyptoton in fair, fairer, fair; in the four-fold repetition of praise (both as a verb and a noun), and the amplification of the numbers created by the rearranged syntax of the final clause: ten times more,/ ten times ten thousand more. Marked literary words also include borrowed words, or, as they were disparagingly called, “inkhorn terms”. The very existence of disparaging terms in the period for the practice of borrowing words itself shows how prevalent the practice became during the Early Modern period and also how the practice was employed: inkhorn terms were born of the inkwell, and therefore (we presume) they were more frequent in literary language than in common speech. Aureate diction (popularized in the 15th century by John Lydgate) continued in the literature of the 16th century and was characterized by its Latinate lexical stylings. Even writers who disapproved of the practice of excessive borrowing from Romance languages were obliged to take issue with it in these same borrowed words because the English lexicon was already so full of loan words. Estimating the borrowing practices using the OED electronic record shows that Latin borrowing exceeded French in the period, and that the influx of Latin loan words peaked between 1575–1675, with more than 13,000 loan words entering the lexicon during that time (Culpeper and Clapham 1996: 218; Nevalainen 2006: 53). It is more difficult to estimate, of course, the difference between bookish usage and oral usage with respect to borrowed words, but it is reasonable to assume that the large-scale Latin borrowing of the Early Modern period was much more characteristic of written, formal, restricted, and technical genres. Since the OED records disproportionately represent literary sources in the Early Modern period, the recorded 13,000 Latin borrowings in the century between 1575–1675 must also be disproportionately literary (see further Lancashire, Chapter 40).
799
800
V Early Modern English Another aspect of literary words is the greater prevalence of compounds, as in (9). This is a long-standing technique of literary style with both classical precedent (the wine-dark sea) and Anglo-Saxon (kennings like whale-road). (9)
a. Saphir-winged Mist (1681 Marvell, Upon Appleton House, to my Lord Fairfax 100) b. dewy-feather’d Sleep (1646 Milton, Il Penseroso 43)
Facilitated by the diminishment of case endings and the more fluid functionality of words, compounding was one of a range of creative solutions that Early Modern English exploited to expand the lexicon. Beattie’s list at the beginning of this section, moreover, suggests that readers perceived compounds to be a particularly characteristic feature of poetic language. Beattie also focuses in his list on the use of abbreviated words. George Gascoigne similarly mentions the practice of reduction as part of poetic licence; he gives examples like “ ‘o’ercome’ for ‘overcome’”, “ ‘ta’en’ for ‘taken’ ”, and “ ‘heav’n’ for ‘heaven’”, but he also mentions additions like “ ‘ydone’ for ‘done’” and substitutions like “ ‘thews’ for ‘good parts’” (Gascoigne 1999 [1575]: 168). Jonathan Swift (1712) complains about this practice as well in his Proposal for Correcting, Improving and Ascertaining the English Tongue. He condemns the practice as “barbarous” and blames Restoration poets for it, claiming that they shorten words for metrical reasons. Yet all of the examples that he gives (drudged, disturbed, rebuked, fledged) have reduced final syllables in Present-day English, and this is probably not owing to the influence of poetry. Swift and Gascoigne’s assessments serve to emphasize that abbreviated words and the use of apostrophes to indicate truncated syllables were a marked feature of poetry, marked to the extent that Swift attempted to blame poetry for the phonological deletion. While “literary style” is certainly composed of elements present in the language as a whole, particular features and constructions are observed to be characteristic of literary genres. In the Early Modern period, these perceptual aspects were likely to be associated with high style or poetic language: classical figures, complex syntax, Latinate words, functional shift, compounding, abbreviation (of words), and amplification (in lexical choice, in phrases, and in discourse). Not all language in literary texts adhered to this description, of course, but the language perceived to be the most literary did.
3 Literary language in its communicative context 3.1 Literature as nationalist project I will now move on to contextualize the ways that this register of literary language assumed a particular cultural function in Early Modern England. The proliferation of literary texts in both the classical and the vernacular languages was accompanied by the rise of literature as an English national project in the Early Modern period. Richard Mulcaster remarks in an oft-cited peroration to his Elementarie: For is it not in dede a meruellous bondage, to becom seruants to one tung for learning sake, the most of our time, with losse of most time, whereas we maie haue the verie same treasur in our own tung, with the gain of most time? our own bearing the ioyfull title of our libertie
50 Early Modern English: Literary language and fredom, the Latin tung remembring us, of our thraldom & bondage? I loue Rome, but London better, I fauor Italie, but England more, I honor the Latin, but I worship the English (Mulcaster 1582: 254).
Mulcaster famously describes here the way that the choice of English for learning and literature is a point of national pride, even a rejection of bondage. Mulcaster, a schoolmaster, presumably passed some of these opinions on to his students, and indeed we find that his most illustrious student, Edmund Spenser, went on to use and adapt the English tongue for his own nationalist projects (as will be discussed below). English language historians have long noted that the developing sense of nationhood in the 16th century was accompanied by greater attention to codifying English. And English literary historians have similarly noted that the 16th century begins a period of increased consciousness of and attention to a national literature. The two of these overlap in the increased attention to codifying literary language in the Early Modern period: the rise in production of rhetorics, studies of poetic language, and guides to figures and schemes. Developing a self-consciously national literature must involve, in part, a consideration of what constitutes literary art. And the proliferation of rhetorical and critical texts testifies to a national energy that developed in the period for classifying and cataloguing literary language. Earlier in the Elementarie, for example, Mulcaster (1582: 77) describes his ambitions: “therefore I will first shew, that there is in our tung great and sufficient stuf for Art: then that there is no such infirmitie in our writing, as is pretended, but that our custom is grown fit to receaue this artificiall frame […]”. Mulcaster asserts that the English language is suitable to be adapted for artistic purposes, and we find a collective initiative in that direction. Explicit attention to the analysis of rhetorical and poetic tools created a larger awareness of them among both writers and readers, making them more fundamental to the expressed characteristics of literary genres and meaning that the act of authorship became in part a dialogue with other crafters of style. The industry of literary creation becomes driven by the need to engage with and personalize figures and tropes, adapting them for the English literary tradition. Puttenham identifies expressly this as his motivation for renaming the classical figures and tropes with English names: But when I consider to what sort of readers I write, and how ill-faring the Greek term would sound in the English ear; then also how short the Latins come to express many of the Greek originals; finally, how well our language serveth to supply the full signification of them both, I have thought it no less lawful, yea, peradventure, under license of the learned, more laudable, to use our own natural, if they be well chosen and of proper signification, than to borrow theirs. So shall not our English poets, though they be to seek of the Greek and Latin languages, lament for lack of knowledge sufficient to the purpose of this art. (Puttenham 2007 [1589]: 3.9.242 [italics mine]).
Puttenham’s manual makes the tools of literary analysis (figures) available to his audience and it brings these to its readers in English rather than by simply importing the Greek and Latin terms. Puttenham’s anglicizations did not catch on, of course, and we now smile at his attempts to rename hyperbole as the Overreacher (3.18.276) or micterismus as the Fleering Frump (3.18.275). But the drive to improve English, to make it a language for art and for analysis was a powerful motivation, and the literary
801
802
V Early Modern English language of the Early Modern period was both a product of and an engine for this national self-improvement impulse. One stylistic feature pressed into nationalistic service was purposefully archaic or conservative English. Archaisms represent for some a purer English, unadulterated by Latin and the romance languages. Edmund Spenser’s The Faerie Queene is typically perceived as the foremost Early Modern example of a poet employing purposeful archaism for nationalist ends, as shown in (10). (10)
High time it seemed then for euery wight Them to betake vnto their kindly rest; Eftsoones long waxen torches weren light, Vnto their bowres to guiden euery guest: Tho when the Britonesse saw all the rest Auoided quite, she gan her selfe despoile, And safe commit to her soft fetherednest, Where through long watch, & late dayes weary toile, She soundly slept, & carefull thoughts did quite assoile (1596 Spenser, The Faerie Queene 407)
Stanzas like the above show Spenser’s stylistic choice of archaic or conservative diction (wight, eftsoones, assoile), his use of old morphosyntactic forms (weren, guiden, gan), and his use of conservative semantic senses (kindly meaning ‘natural’). The conservative forms and lexical choices evoke for the reader the early English poetic tradition (wight, for example, is a common word in Middle English alliterative poetry). The use of these older forms and words created a veneer of older language for readers of The Faerie Queene, instantiating through the style the creation of an authorizing epic for Elizabeth I, and lending it an air of solemnity and historicity. To write an English epic, then, Spenser selects what he sees as the purest “English” in the linguistic repertoire. The dedicatory epistle to his Shepheardes Calender (Spenser 1579) sets out the philosophy of such literary projects. This prefatory section (by an author named as “E. K.”, but who may be Spenser himself) sets out Spenser’s ambitions to restore older English: he hath laboured to restore, as to theyr rightfull heritage such good and naturall English words, as have ben long time out of use and almost cleare disherited. Which is the onely cause, that our Mother tonge, which truely of it self is both ful enough for prose and stately enough for verse, hath long time ben counted most bare and barrein of both. Which default when as some endevoured to salve and recure, they patched up the holes with peces and rages of other languages, borrowing here of the french, there of the Italian, every where of the Latine, not weighing how il, those tongues accorde with themselves, but much worse with ours: So now they have made our English tongue, a gallimaufray or hodgepodge of al other speches (1579 Spenser sig. ¶2v).
E. K. praises Spenser for reviving long-neglected English words to remedy the deficits perceived in English as a literary language. He applauds the use of archaism over large-scale borrowing. In this ideological framework, older words were commonly regarded as better and more venerable, and newer words (especially borrowed words) as suspect.
50 Early Modern English: Literary language Archaism can come in the choice of Anglo-Saxon words, or in the form of older morphosyntactic variants. The inflectional ending -s for the third person present singular verb appears with more and more frequency in the 16th and 17th centuries in the place of the -th endings (he goes rather than he goeth), but the Early Modern period witnesses the coexistence of the two: the -s ending is the more progressive form and the -th ending is the more conservative (Nevalainen 2006: 89–92; see further Cowie, Chapter 38). This permits a choice between the two: a choice which could be partly syntactic. Frequency is one predictor for the use of more conservative forms (thus, the -th forms doth, hath and saith are more resistant than most other verbs), but so is the genre of the text: the higher the style of the writing, the more conservative the forms are likely to be. The Authorized Version of the Bible (the King James version), for example, is well-known for immortalizing some of these older morphosyntactic variables as religious-sounding language by consistently employing the conservative -th forms, as well as now-archaic plurals like kine and brethren (see Kohnen, Chapter 65). The desire of literary language to utilize styles that are elevated and formal has kept variants alive in poetry long after they were moribund in speech. Consider the use of the second person thou/thee pronouns which were utilized in poetry even through the 19th-century, and still sound vaguely poetic to our contemporary ear (see further Busse, Chapter 46). Examples like these second person singular pronouns show us how morphosyntactic features can move from functioning grammatically and referentially to functioning also as markers of a literary register. Early Modern texts do show some overlap between the purposeful use of archaism and the literary use of dialect. It can often be difficult to distinguish dialectal forms from archaic ones; the use of Northern dialectal forms in poems like The Shepheardes Calender (Spenser 1579) creates confusion in many places about whether Spenser was trying to employ a dialectal form or an archaic one. The two, indeed, were viewed as related: Northern English was seen as older and less tainted by foreign borrowings (Blank 1996: 100). The dialect used in poetry was apparently usually the English of northern England, if Alexander Gil is to be believed. Northern English attained the stylistic gravitas of an acceptable regional dialect for poetry, in a way that the dialect of, for example, Dorset only did later.
3.2 Language play The use of dialect also functions in this period as comic material, especially in Early Modern drama: this period begins to see more evidence of dialect comedy and playful contrasts in register. Although literary depiction of dialect does not originate in the Renaissance – we see some dialect representation in literary language in Chaucer, for example – it becomes much more developed, and there is more evidence of dialect as a stylistic tool for literary use, for characterization, realism, disguise, comedy. This includes the use of regional dialects and also some subcultural dialects, such as cant, the language of the London underworld. The specialized lexicon of thieves and other lawbreakers was apparently of great interest to the reading public – over 100 glossaries of cant and slang were published between 1567 and 1784, according to Coleman (2008). Cant had quite a developed literary presence, and a number of plays feature the use of cant: Thomas Dekker and Thomas Middleton’s The Roaring Girle (1607–1610), Richard Brome’s A Jovial Crew or the Merry Beggars (1641), and Francis Beaumont, John Fletcher, and Philip Massinger’s
803
804
V Early Modern English The Beggars’ Bush (1622). The literature of cant also includes other descriptions of social outliers, from Samuel Rowlands’s Slang Beggars’ Songs (1610) to Daniel Defoe’s The Complete Mendicant (1699) (all available – with the exception of Rowlands – on Early English Books Online [Chadwyck-Healey 2003–2011]). The Early Modern period employed language for other playful purposes: flyting (the art of insulting), political satire, punning. Some texts accomplish all of the ends at once – more than one of John Skelton’s verses is devoted to ridiculing assorted public figures, as in (11). (11)
His hed is so fat He wotteth neuer what Nor wherof he speketh He cryeth and he creketh He pryeth and he peketh He chydes and he chatters He prates and he patters He clytters and he clatters He medles and he smatters He gloses and he flatters Or yf he speake playne Than he lacketh brayn (1545 Skelton, Colyn Cloute A2v)
This passage from Colyn Cloute shows a combination of alliteration, rhyme, and onomatopoeia that collude for the purposes of ridicule. And studying the aesthetic pleasures of language entails studying the aesthetic pleasures of mockery. From comic plays to humorous verse, literature gives language some of its most playful genres. It even does so with literary language as the butt of the joke; to the extent that the high style provided models for elevated tones in literary writing, it also provided parallel fodder for mockery and lampoon. Characters like Holofernes in Shakespeare’s Love’s Labour’s Lost serve as exaggerated users of high style features, and show the pomp and preciousness of Renaissance ornate styling. There is even a name for mock-ridiculous high style: fustian style. In 1599 John Hoskyns (cited in Adamson 1999: 576) spoke with pride about his “fustian speech” – a speech delivered in the Middle Temple. Adamson points out that both the terms bombast (the overuse of Latinisms) and fustian (their playful or anarchic use) develop their metalinguistic senses at the end of the 16th century – both developing out of words for clothing (Adamson 1999: 576). Making fun of the figures and tropes of elevated writing was a way of deflating some of the “height” of high style, but it also serves as evidence for how pervasive these stylistic techniques were. One kind of linguistic joke about the misuse of language came through humorous speech errors on the part of comic characters. Although this comic use of misspeaking is not named “malapropism” until Sheridan’s Mrs. Malaprop lends her name to the practice in 1775, the literary exploitation of language errors is, of course, much older. Peacham (1577: sigs. G2v–G3r) calls the figure cacozelon and the Renaissance provides several famous examples: Shakespeare’s Mistress Quickly and Dogberry among them. The category of language play might even extend to include aspects of literary language utilized for aesthetic pleasure, such as the rise of certain metrical forms. Metrical
50 Early Modern English: Literary language structure is a feature of form, certainly, but it also becomes an aspect of cultural context. Communities of writers develop and popularize particular forms, and these forms attain significance within this cultural setting. I will not give here a history of the cultural significance of Early Modern metrics, but it is notable that the Renaissance saw an explosion of energies in metrical form. The period saw such major innovations, for example, as the importing of the sonnet by Sir Thomas Wyatt in the early 16th century, a form that quickly established itself as central to English poetry. As English poets explored the capability of English as a language for poetry, they also attempted to adapt foreign metrical forms to suit English. Some ambitious attempts were ultimately regarded as unsuccessful (see Attridge’s 1974 book about the many poets who tried to make the classical dactylic hexameter line work in English), but the shift in poetic meter, the development of the Spenserian stanza, and the development of the blank verse form were radical transformations for the literary tradition. Blank verse, in particular, demonstrated a breadth and flexibility that have made it one of the predominant poetic forms in English. Developments and fads in poetic form are revealing about the ideologies and influences of literary language: how poetry is made English and how the English made poetry.
4 Summary Literary language is language with a specialized pragmatic purpose, and to examine the literary language of the Early Modern period is to provide perspective on the English that pleased and amused 16th- and 17th-century speakers. It is also the medium for the narratives that people construct about their lives. This is seen in the Early Modern period, for example, in the nationalist pursuit of a particularly English literary tradition. Literary language functions pragmatically in such ways as a culture has defined its notions of literature. In the Renaissance, literary texts – and by extension literary language – were taken as didactic, entertaining, beautiful, morally edifying, and culturally unifying. The stylistic elements that heightened or marked these effects include the decorative elaborations, the stylized comparisons, the elegant repetitions, the passionate metaphors, the comic manglings, and the meticulous figures. All indicate how speakers and writers crafted their language for aesthetic purposes. Certainly it is a testament to their success that we still read and appreciate Early Modern literary texts from many centuries’ remove.
5 References Adamson, Sylvia. 1999. The literary language. In: Lass (ed.), 539–653. Adamson, Sylvia. 2001. The grand style. In: Sylvia Adamson, Lynette Hunter, Lynne Magnusson, Ann Thompson, and Katie Wales (eds.), Reading Shakespeare’s Dramatic Language: A Guide, 31–50. London: Arden Shakespeare. Adamson, Sylvia. 2007. Synonymia: or, in other words. In: Adamson et al. (eds.), 17–36. Adamson, Sylvia, Gavin Alexander, and Katrin Ettenhuber (eds.). 2007. Renaissance Figures of Speech. Cambridge: Cambridge University Press. Attridge, Derek. 1974. Well-Weigh’d Syllables: Elizabethan Verse in Classical Metres. Cambridge: Cambridge University Press. Blank, Paula. 1996. Broken English: Dialects and the Politics of Language in Renaissance Writings. London: Routledge.
805
806
V Early Modern English Carter, Ronald and Walter Nash. 1983. Language and literariness. Prose Studies 6(2): 123–141. Chadwyck-Healey. 2003–11. Early English Books Online, 1475–1700 (EEBO). Ann Arbor: ProQuest. http://eebo.chadwyck.com/home Coleman, Julie. 2008. A History of Cant and Slang Dictionaries. Oxford: Oxford University Press. Culpeper, Jonathan and Phoebe Clapham. 1996. The borrowing of Classical and Romance words into English: A study based on the electronic Oxford English Dictionary. International Journal of Corpus Linguistics 1(2): 199–218. Donne, John. 1633. Poems by J.D. With elegies on the authors death. Early English Books Online. Erasmus, Desiderius 1963 [1512]. On Copia of Words and Ideas. Trans. by Donald B. King and H. David Rix. Milwaukee: Marquette University Press. Fraunce, Abraham. 1588. The Arcadian rhetorike, or The praecepts of rhetorike made plaine by examples Greeke, Latin, English, Italian, French, Spanish … London: Printed by Thomas Orwin. Early English Books Online. Gascoigne, George. 1999 [1575]. A primer of English poetry. In: Vickers (ed.), 162–171. Gil, Alexander. 1621. Logonomia Anglica. Series Logonomia Anglica. (English Linguistics 1500–1800, ed. by R. C. Alston, 68.) Menston: The Scolar Press, 1968. Gilbert, A. J. 1979. Literary Language from Chaucer to Dryden. London: Macmillan. Go¨rlach, Manfred. 1991. Introduction to Early Modern English. Cambridge: Cambridge University Press. Herbert, Mary (Sidney), Countess of Pembroke 1823 [c.1599]. Psalm 52. In: Philip Sidney and Mary (Sidney) Herbert, Pembroke, The Psalmes of David translated into Divers and Sundry Kindes of Verse, more rare and excellent for the method and varietie than ever yet hath been done in English. Now first printed from a copy of the Original Manuscript, transcribed by John Davies of Hereford in the reign of James the First, 91–93. London: Chiswick Press. Herman, Vimala. 1983. Introduction: Special Issue on Literariness and Linguistics. Prose Studies 6(2): 99–122. Hoskins, John. 1935 [1599]. Directions for Speech and Style. Ed. by Hoyt Hudson. Princeton: Princeton University Press. Hoskyns, John. 1999 [1599]. Sidney’s Arcadia and the rhetoric of English prose. In: Vickers (ed.), 398–427. Jonson, Ben. 1640. Timber: or Discoveries; Made Vpon Men and Matter: As they have flow’d out of his daily Readings; or had their refluxe to his peculiar Notion of the Times. In: The workes of Benjamin Jonson. The Second Volume. Containing These Playes viz. 1. Bartholomew Fayre. 2. The Staple of Newes. 3. The Divell is an Asse, 410–435. London: Printed for Richard Meighew. Early English Books Online. Jonson, Ben. 1999 [1615–35]. Notes on Literature. In: Vickers (ed.), 558–589 Lamy, Bernard. 1676. The art of speaking written in French by Messieurs du Port Royal in pursuance of a former treatise intituled, The art of thinking. London: Printed by W. Godbid to be sold by M. Pitt. Early English Books Online. Lass, Roger (ed.). 1999. Cambridge History of the English Language. Vol. III. 1476–1776. Cambridge: Cambridge University Press. Lyly, John. 1578. Euphues. The anatomy of vvyt Very pleasant for all gentlemen to reade, and most necessary to remember: wherin are contained the delights that wyt followeth in his youth, by the pleasauntnesse of loue, and the happynesse he reapeth in age, by the perfectnesse of wisedome. London: T. East for Gabriel Cawood. Early English Books Online. Marlowe, Christopher. 1598. Hero and Leander: begun by Christopher Marloe; and finished by George Chapman. 2nd edn. London: Printed by Felix Kingston, for Paule Linley. Early English Books Online. Marvell, Andrew. 1681. Upon Appleton House, to my Lord Fairfax. In: Miscellaneous Poems. London: Printed for Robert Boulter. Early English Books Online. Milton, John. 1646. Il Penseroso. In: Poems of Mr. John Milton, both English and Latin, compos’d at several times. Printed by his true copies. / The songs were set in musick by Mr. Henry Lawes
50 Early Modern English: Literary language
807
Gentleman of the Kings Chappel, and one of His Maiesties private musick. Printed and publish’d according to order. London: Printed by Ruth Raworth for Humphrey Moseley. Early English Books Online. Milton, John. 1667. Paradise Lost. A poem written in ten books. London: Printed and are to be sold by Peter Parker.., and Robert Boulter…, and Matthias Walker. Early English Books Online. More, Henry. 1642. Psychodia platonica, or, A platonicall song of the soul consisting of foure severall poems … : hereto is added a paraphrasticall interpretation of the answer of Apollo consulted by Amelius, about Plotinus soul departed this life. Cambridge: Printed by Roger Daniel, printer to the Universitie. Early English Books Online. Mulcaster, Richard. 1582. The First Part of the Elementary. (English Linguistics 1500–1800, ed. by R. C. Alston, 219.) Menston: The Scolar Press, 1970. Nevalainen, Terttu. 2006. An Introduction to Early Modern English. Oxford: Oxford University Press. North, Dudley. 1645. A Forest of Varieties. London: Printed by Richard Cotes. Early English Books Online. Partridge, A. C. 1971. The Language of Renaissance Poetry: Spenser, Shakespeare, Donne, Milton. London: Andre Deutsch. Peacham, Henry. 1577. The garden of eloquence conteyning the figures of grammer and rhetorick, from whence maye bee gathered all manner of flowers, coulors, ornaments, exornations, formes and fashions of speech, very profitable for all those that be studious of eloquence, and that reade most eloquent poets and orators, and also helpeth much for the better vnderstanding of the holy Scriptures. London: H. Iackson. Early English Books Online. Puttenham, George. 2007 [1589]. The Art of English Poesy. Ed. by Frank Whigham and Wayne A. Rebhorn. Ithaca: Cornell University Press. Rissanen, Matti. 1999. Syntax. In: Lass (ed.), 187–326. Shakespeare, William(?). 1596. The Raigne of King Edvvard the third: As it hath bin sundrie times plaied about the Citie of London. London: Printed for Cuthbert Burby. Early English Books Online. Sherbo, Arthur. 1975. English Poetic Diction from Chaucer to Wordsworth. East Lansing: Michigan State University Press. Sidney, Philip. 1970 [1595]. An Apology for Poetry. Indianapolis: Bobbs-Merrill. Skelton, John. 1545. Here after foloweth a litel boke called Colyn Cloute compyled by mayster Skelton poete Laureate. 2nd edn. Early English Books Online. Spenser, Edmund. 1579. The shepheardes calender conteyning twelue aeglogues proportionable to the twelue monethes. Entitled to the noble and vertuous gentleman most worthy of all titles both of learning and cheualrie M. Philip Sidney. London: Printed by Hugh Singleton. Early English Books Online. Spenser, Edmund. 1596. The Faerie Queene Disposed into twelue bookes, fashioning XII. morall vertues. 2nd edn. London: printed by Richard Field for VVilliam Ponsonbie. Early English Books Online. Swift, Jonathan. 1712. Proposal for Correcting, Improving and Ascertaining the English Tongue. (English Linguistics 1500–1800, ed. by R.C. Alston, 213.) Menston: The Scolar Press, 1969. Tynyanov, Yury. 2000 [1924]. The literary fact. In: David Duff (ed.), Modern Genre Theory, 29–49. Harlow: Pearson Education. Vickers, Brian (ed.). 1999. English Renaissance Literary Criticism. Oxford: Clarendon Press. Wilson, Thomas. 1909 [1560]. Arte of Rhetorique. Ed. by G. H. Mair. Oxford: Clarendon Press. Wroth, Lady Mary. 1621. The Countesse of Mountgomeries Urania. London: Printed for Iohn Marriott and Iohn Grismand. Early English Books Online.
Colette Moore, Seattle (USA)
808
V Early Modern English
51 Early Modern English: The language of Shakespeare 1 2 3 4 5 6 7 8
Introduction General considerations Phonology Vocabulary Grammar Pragmatics and discourse Recent trends and further directions References
Abstract The present chapter provides a brief outline of Shakespeare’s linguistic contribution to EModE. As a starting point, important reference works and tools for further research are introduced. Then some general considerations on Shakespeare’s language are presented, because Shakespeare’s language, or rather his use of language can, and in fact has been defined in broader or narrower terms. The description and illustration of Shakespeare’s language begins with a section on pronunciation, then moves on to vocabulary and grammar and ends with pragmatic and sociolinguistic studies and other relatively recent trends. Each section begins by listing basic works on a topic and then presents more detailed or more advanced studies, usually in chronological order. Where necessary, illustrative examples are provided.
1 Introduction For the final decades of the Renaissance the works of William Shakespeare (1564–1616) and the King James Bible (the Authorized Version) of 1611 are the dominating influences: “Dominate, that is, from a linguistic point of view. The question of their literary brilliance and significance is not an issue for this book. Our question is much simpler yet more far-reaching: what was their effect on the language?” (Crystal 1988: 196). Shakespeare’s linguistic quality can be seen on many levels of language. Insightful observations have turned proverbial, one among those is that love is blind. Some Shakespearean words, such as powerfully or obscenely, are still used today, others, such as indirection or incarnadine, are no longer used, but they seem to have a particularly challenging contextual quality. Generally speaking, the number of words Shakespeare invented is impressive (see Section 4). His word-stock is equally striking. Depending upon one’s definition of word, different results concerning the total size of the Shakespeare corpus may be obtained. If the word count of the Spevack concordances (1968–80) is taken as a reliable basis, the sum total of words used in the 38 plays and his non-dramatic works amounts to 884,647 (see Spevack 1968–80 IV [1969]: 1). Spevack also supports the importance of Shakespeare’s English when he states that “indeed, our picture of English as a whole will be improved by a detailed study of all of Shakespeare’s language not only because Shakespeare, we will agree, may be the greatest practitioner of English but certainly because he accounts for about 40 per cent of the recorded English of his time” (Spevack 1972: 108). This outstanding position is also reflected in the table of contents of the present volume because Chaucer Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 808–826
51 Early Modern English: The language of Shakespeare (see Horobin, Chapter 36) and Shakespeare are the only authors to whom separate chapters are devoted.
1.1 Reference works and research tools Linguistic contributions to Shakespeare’s English date back at least to the tradition of late 19th-century scholarship. There we find grammars and dictionaries that list and describe linguistic peculiarities of Shakespeare’s language use. The first comprehensive account is Abbott’s (1966 [1870]) Shakespearian Grammar; Franz (1986 [1898/99, 1939]) follows in this tradition. In 1990, Blake (1990: 61) deplored that “we still badly need a grammar on modern principles and a new comprehensive dictionary of Shakespeare’s language”. With two new Shakespeare grammars, one written by Blake (2002) himself, and another one by Hope (2003), this wish finally came true, at least for grammars. It is perhaps not so incidental that the two grammars came out at this very point in time, because both reflect a revived interest of modern linguistics in historical language studies. In the introductory chapters of their books, both Hope and Blake point out that for more than 130 years the grammar by Abbott has been the standard work for the English-speaking world, although those scholars familiar with German could resort to Wilhelm Franz’s 1986 [1898/99, 1939] Shakespeare-Grammatik. Blake argues that “there have been no substantial grammars of Shakespeare’s language since Franz’s, and that the grammars that have appeared are altogether slimmer volumes with relatively little detail. In addition, most of them have built upon the work of Franz and Abbott (e.g. Brook 1976 and Scheler 1982)” (Blake 2002: 10). Both modern grammars take a descriptive approach, but the scope of Blake’s book is wider than that of Hope’s. While Hope covers grammatical ground up to the clause and sentence level, Blake in addition to this includes two chapters on discourse and register and on pragmatics. For a detailed review of both grammars see U. Busse (2007). The oldest Shakespeare dictionaries also date from the late 19th century. Schmidt and Sarrazin (1971 [1874/75]) still provide valuable insights, as all of the entries are amply illustrated by quotations. Less comprehensive in that they concentrate on difficult words are the dictionaries by Cunliffe (1910), Onions (1911), and Kellner (1922). In contrast to grammar writing, we still lack a modern comprehensive Shakespeare dictionary, but we have got a number of modern specialized dictionaries, thesauri, and glossaries. Most importantly, there is Spevack’s (1993) Shakespeare Thesaurus. Based on a historically informed semantic categorization of the Shakespeare lexicon this glossary contains 15 semantic categories ranging from animal terms to body and bodily functions to plants, religion, and time. Crystal and Crystal (2002) is in parts close to Spevack (1993). Leisi (1997) deals with difficult or problematic words and passages in Shakespeare’s plays. There are also specialized dictionaries, such as those that focus on Shakespeare’s use of sexual language (Partridge 1955; Rubinstein 1984; Williams 1997), or on legal language (Sokol and Sokol 2000). The nine volumes of Spevack’s (1968–80) A Complete and Systematic Concordance to the Works of Shakespeare are based on Evans’s (1972) Riverside Shakespeare. This concordance supersedes the older one by Bartlett (1894). The nine volumes are an unprecedented achievement that goes beyond the usual requirements of a concordance in that they provide the frequencies and words used by each character in the respective
809
810
V Early Modern English plays. Volumes 1–3 focus on words used in the comedies, histories and tragedies and follow the generic distinction of the First Folio (F1) from 1623. Volumes 4 and 5 provide the words of all the plays and poems in alphabetical order. Volume 6 contains the appendices and special cases, such as hyphenated words, hapax legomena, and so on, and Volume 7 is a concordance to the stage directions and speech prefixes. Volume 8 is a concordance to the bad quartos, and Volume 9 is a concordance to the substantive variants. Perhaps the most comprehensive current investigation into Shakespeare’s language is the Shakespeare Database carried out at the University of Mu¨nster in Germany under the direction of H. Joachim Neuhaus (forthc.). The database constitutes a full lemmatization of Shakespeare’s work, which is organized into a relational database and follows Evans’s (1997 [1972]) Riverside Shakespeare. Unfortunately, this database is still not publicly available. Wordhoard (Parod 2004–2010), which is a joint project of the Perseus Project at Tufts University and The Northwestern University Library, presents among other texts like Homer’s The Odyssey, Chaucer’s The Canterbury Tales and Edmund Spenser’s works, a lemmatized and morphosyntactically tagged database of Shakespeare’s language. It also uses a modern edition of Shakespeare. Regarding bibliographies, there are Bradbrook’s (1954) retrospect “Fifty years of the criticism of Shakespeare’s style” and Partridge’s (1979) “Shakespeare’s English: A bibliographical survey” as older works. Adamson et al. (2001: 302–315) provide an annotated bibliography of the most important monographs and collections of Shakespeare’s English, ranging from Ko¨keritz (1953) to the new millennium, and concentrating on works that appeared in the 1980s and 1990s. The comprehensive bibliography in Blake’s (2002: 346–361) grammar lists titles ranging from 1870 (Abbott) to 2000 (Kermode). Apart from the reference works and research tools above, a number of book-length accounts on Shakespeare’s English appeared from the 1960s to the early 1980s, as, for instance, Hulme (1962, 1972), Brook (1976), Hussey (1982), Scheler (1982), and Blake (1983). In addition to these monographs, there are the collections by Muir and Schoenbaum (1971) and by Salmon and Burness (1987), which contains thirty-three essays published between 1951 and 1983. The studies cover the following subject areas: “Shakespeare and the English language”, “Aspects of colloquial Elizabethan English”, “Studies in vocabulary”, “Shakespeare and Elizabethan grammar”, “Studies in rhetoric and metre”, “Punctuation”, and “The linguistic context of Shakespearean drama”. After many years during which linguistic descriptions of Shakespeare were not high on the agenda of historical linguists, the two grammars together with other recent publications such as Crystal (2008) show a renewed interest in the subject. Some of these, e.g. Adamson et al. (2001), Kermode (2000), Magnusson (1999), and McDonald (2001), combine linguistic investigations with aspects of rhetoric and style. For example, Adamson et al. (2001) is directed both at an academic readership and at teachers and students. Therefore, the collection takes on an interdisciplinary perspective and the chapters range from those that focus on rhetoric and style, to Shakespeare’s word-formation and his use of puns to techniques of persuasion.
2 General considerations In a classical article, Quirk makes the following observations: “It should be superfluous to point out that the language of Shakespeare is an amalgam of the language that
51 Early Modern English: The language of Shakespeare Shakespeare found around him – together with what he made of it. […] It is necessary, therefore, to study the language of Shakespeare’s time and then to distinguish Shakespeare’s language within it” (Quirk 1987: 3, 4). In order to do so, Quirk proposes a threefold distinction: 1. English as it was about 1600 2. Shakespeare’s interest in his language 3. Shakespeare’ unique use of English (Quirk 1987: 4). However, the 400 years of linguistic change on all levels of the language system and in language use, will leave a “discrepancy between Shakespeare’s intuitions about language and our own” (Crystal 2003: 67). Although we have to be aware of the fact that our modern intuitions may lead us astray, a “modern historical linguistics” (Mair 2006) approach, encompassing socio-historical, pragmatic, cognitive, and corpus-linguistic aspects, makes it possible to reconstruct the past and to study the language of Shakespeare from new angles. Methodologically, a number of different approaches with different objectives or foci are possible. Salmon (1987: xiii) lists four different approaches to Shakespeare’s English, which are exemplified by the essays collected in Salmon and Burness (1987). The four groups of essays pursue the following purposes: 1. to enable the speaker of Present-day English to share as far as possible in the responses of the original audience; 2. to draw attention to the manner in which Shakespeare handles the language of his time for artistic purposes; 3. to use the language of Shakespearean drama as data on which to base conclusions about Elizabethan English in general; and 4. to provide illuminating information about various aspects of Shakespeare’s linguistic background, in particular, the attitude to the language of his time. To any of these ends, the object of investigation – Shakespeare’s language – can be defined in narrower or broader terms. For instance, Blake (1990: 63) argues that “the language of Shakespeare has received little attention in the twentieth century – at least in so far as language is concerned with the grammatical system of expression”. In this context, he comments on Evans’s (1952) The Language of Shakespeare’s Plays, arguing that “[t]his book says nothing about the grammar and little about the vocabulary of the plays. The Language in the title means either style or comments by the characters on language” (Blake 1990: 63). However, at present, the formerly distinct research traditions of linguistics and stylistic or literary studies are coming into closer contact also within Shakespeare studies under what one could label “the new philology” (see Taavitsainen and Fitzmaurice 2007: 22–25). As recent examples for this direction of research, one could take the work by Magnusson (1999, 2007), Adamson et al. (2001), Spevack (2002), or B. Busse (2006b). Regardless of the purpose or scope of a given study, investigating Shakespeare’s language always demands attention to those issues that are generally relevant to the analysis of historical data. These mainly revolve around the questions of how historical linguists (need to) validate their data and analyses. It also includes a consideration of such issues as the edition on which an investigation is based (e.g. modern edition or
811
812
V Early Modern English early printings). Diachronic investigations of Shakespeare’s language that focus on language change and stability or a linguistic analysis that takes account of the respective genres need to meticulously consider the specifically Shakespearean textual and editorial conditions and the ways some modern categorizations deviate from historical ones. For example, the First Folio from 1623 only distinguishes between comedies, histories, and tragedies and does not contain the genre label “romance,” because it is a modern category. Also, it does not contain any act, scene, or line divisions. The extent to which Shakespeare’s language can or cannot be regarded as a true representation of authentic spoken EModE has been frequently discussed (Salmon 1987 [1965]). Within this discussion the fictional character as well as the verse and prose distinction needs to be observed, on the one hand, while, on the other, it can be assumed that Shakespeare made use of the language around him and exploited it creatively for his dramatic purposes (see also Moore, Chapter 50).
3 Phonology Brook (1976: 140–159) and Scheler (1982: 17–32) provide short and practical introductions to Shakespeare’s pronunciation, spelling, and punctuation, and Lass (2001) offers a recent and short chapter on “Shakespeare’s sounds”. He points out that “[m]odern readers or playgoers are auditorily misled by their experience of Shakespeare”. This can be attributed to the fact that we “read Shakespeare in editions with modern spelling and punctuation, and hear him performed with modern English pronunciation” (Lass 2001: 256). Thus, changes in vowel quantity or quality can obscure rhymes, and word stress may occur at unexpected positions. For instance, in order to work out puns, it is essential to know about the homophony of the words and their differing sense. To illustrate the major differences between EModE pronunciation and that of standard PDE, Lass provides the following passage, (1), from A Midsummer Night’s Dream: (1)
Through the forest have I gone; But Athenian found I none On whose eyes I mi(gh)t approve This flower’s fo(r)ce in stirring love. Ni(gh)t and silence – Who is here? Weeds of Athens he doth wear: This is he my master said Despis(e)d the Athenian maid; And he(r)e the maiden, sleeping sound, On the dank and di(r)ty ground. (Midsummer Night’s Dream II.ii.65–74; Lass 2001: 267)
In this passage “[v]owels that have undergone major change in pronunciation are in bold [straight characters]; rhymes which were possible in Shakespeare’s time but are no longer so are in bold italics”. Vowels or consonants that are no longer pronounced are in round brackets. The rs are marked as well, because many dialects do not now pronounce them in these positions (Lass 2001: 267). In addition to this, EModE pronunciation was not yet codified, and therefore much more variable, implying also a greater availability of dialect forms.
51 Early Modern English: The language of Shakespeare Lass (2001: 268) deplores that “[t]here is unfortunately no comprehensive, elementary introduction to Shakespeare’s pronunciation”, but he recommends Ko¨keritz (1953) and his own chapter in the Cambridge History of the English Language (Lass 1999: 56–186), which, he himself describes as “fairly technical, and not devoted to Shakespeare” (Lass 2001: 268), but covering the period as a whole. The assessment of EModE pronunciation can be controversial. While Cercignani (1981), Scheler (1982: 31), and Barber (1997: 140–141) are critical about Ko¨keritz from a scholarly point of view, Adamson et al. (2001: 306–307) stress its usefulness for the purpose of students. According to Barber (1997: 140), the two volumes of Dobson (1968) can be considered as “[t]he standard work on English phonology in the period” in a “very conservative” representation (see Schlu¨ter, Chapter 37). Recently, two books dealing with EModE and Shakespearean sound appeared, namely Smith (1999) and Folkerth (2002). The Globe Theatre now offers selected performances of Shakespeare’s plays in EModE pronunciation. Crystal (2005) reports on this project (audio extracts are available at The Shakespeare Portal [Crystal 2005]). The success of the project calls into question an earlier position, as pronounced by Quirk in 1971: Now that we have the technical ability to put on a play in roughly the pronunciation of 1600, the desirability of so doing has become less apparent. Since so many of the features of Elizabethan pronunciation have remained in twentieth-century use with utterly different sociological connotations, it is exceedingly difficult to avoid farcical overtones (Quirk 1987: 5 [1971: 69, 1974: 48]).
4 Vocabulary In the introduction to his Shakespeare grammar, Blake (2002: 7) says that “[t]he language of Shakespeare has been studied and his texts edited almost from the moment he died in 1616” and that two features have remained constant factors of scholarship and criticism: vocabulary and grammar. It has been claimed that the EMod period, and in particular the time span from 1530 until 1660 saw an unprecedented growth in vocabulary (see Lancashire, Chapter 40). For example, on the basis of the Chronological English Dictionary (Finkenstaedt et al. 1970), 3,300 neologisms are attested for the year 1600 alone (see Go¨rlach 1994: 109). As concerns the means of expanding the lexicon, “substantial use was made both of borrowing (especially from Latin) and of word-formation (especially by affixation)” (Barber 1997: 219). The so-called “inkhorn controversy” discussed the desirability of vocabulary growth and its appropriate methods. These topics are also reflected in Shakespeare’s English. People talk a lot about Shakespeare’s “linguistic legacy”, saying that he was a major influence on the present-day English language, and citing as evidence his coining of new words (such as assassination and courtship) and idiomatic phrases (such as salad days and cold comfort). But when we add all of the coinages up, we do not get very large numbers. No-one has carried out a precise calculation (Crystal 2003: 77).
This statement is not quite true, as a number of estimations have been made. Scheler (1982: 89–90) lists the following approaches. As concerns the overall size of
813
814
V Early Modern English Shakespeare’s vocabulary, Brook (1976: 26) estimates it to be “of the order of twenty thousand words”. Spevack (1968–80 [1969], IV: 1) counts 29,066 different words (i.e. lemmas). His count is based on the New Riverside Shakespeare and counts different word forms (tokens) such as cried, criedst, cries, and not types or lexemes. Scheler, on the other hand, carries out a type-based count and uses the Shakespeare-Lexicon of Schmidt and Sarrazin (1971 [1874/1875]) as a basis, coming to a total of 17,750 words. However, these statistics do not tell us anything about the number of words that are coined by Shakespeare. The Shakespeare Database (see Section 1.1) lists 3,179 lemmata (4,512 word forms) coined by Shakespeare (B. Busse 2006b). Despite the difficulty that a word first attested in Shakespeare’s work need not necessarily be an original coinage, Scha¨fer (1973: 204–220) lists 1,630 words by year. His list of 308 hyphenated words is far from complete. For the methodological difficulties in tracking first quotations with the help of the OED see Scha¨fer (1980). In terms of etymology (meaning the immediate donor language and not the ultimate source language), Shakespeare’s word stock consists of 43% Germanic words and 54% non-Germanic words (mostly Romance-Latin), if only types are counted, another 3% are difficult to classify (see Scheler 1982: 90). On “Shakespeare’s Latinate Neologisms” and on “Latin-Saxon Hybrids in Shakespeare and the Bible” see the two articles by Garner (1987a, b). For a more recent comprehensive study of the etymology of Shakespeare’s vocabulary see Franken (1995). Shakespeare systematically exploits the stylistic difference between Germanic and Romance vocabulary. According to Romaine (1982: 122) “the working principle of sociolinguistic reconstruction must be the ‘uniformitarian principle’ ”, which implies that, as in PDE, the Germanic word is the ordinary word, often being more personal, more lively, or more affectionate. By contrast, the Romance loanword is more unpersonal, sought-after, or more dignified. This distinction, both then and now, presupposes from the language users a kind of implicit etymological knowledge, language awareness, or at least a feel for the language. Elizabethans were keenly aware of linguistic decorum and appropriate style (see Adamson 2001a, which includes annotated references for further reading). Apart from decorum, “copia”, which could be expressed by Germanic-Romance synonym pairs, was another important stylistic device. Scha¨fer (1973: 184–203 [appendix II]) lists 529 of these synonym pairs. In order to highlight the stylistic potential of such pairs, Scheler discusses the following example from Macbeth, (2): (2)
Will all great Neptunes Ocean wash this blood Cleane from my Hand? no: this my Hand will rather The multitudinous Seas incarnardine [sic!], Making the Green one, Red (Macbeth II.ii.60–63; Scheler 1982: 93)
The atrocity of Duncan’s murder, and Macbeth’s guilt is expressed in a hyperbole, whose force is amplified by the contrast between the Romance incarnadine and the plain Germanic make red. Inappropriate use of difficult Romance vocabulary by uneducated speakers is a constant source of linguistic humor in the Elizabethan theater. Especially the idiolect of speakers representing the lower social orders has attracted some scholarly attention.
51 Early Modern English: The language of Shakespeare See, for instance, Scha¨fer (1973: 97–115), Brook (1976: Chapter 9, which includes dialects, registers and idiolects), Scheler (1982: 97–100), or Blake (1983: Chapter 2 varieties, Chapter 3 vocabulary, 2002: Chapter 8, 300–302). Schlauch (1987 [1965]) reports on Shakespeare’s characters’ use of malapropisms. With idiolectal language use, the author slips into a character, and his or her linguistic peculiarities can serve the purpose of characterization. This does not only go for uneducated characters such as Mistress Quickly, who, for example, mixes up honeysuckle for ‘homicidal’, honey-seed for ‘homicide’ (2 Henry IV, II.i.47, see Quirk (1987: 14) for this and further examples), but also for characters such as Polonius (in Hamlet) making use of affected, verbose, and bombastic language. In Love’s Labors Lost, Holofernes and Nathaniel are laughed at for having “been at a great feast of languages and stol’n the scraps” (V.i.34). Examples such as these do however not imply that Shakespeare was against augmentation of vocabulary and scorned inkhorn terms, but that he wanted to draw attention to the fact that “the uneducated would make ridiculous errors” and others would be ridiculous because of their “use of learned language for obscurity’s sake” (Quirk 1987: 14). Since dialogue is the most important constitutive device of drama, the playwright pays particular attention to mark socio-regional variants (see Meurman-Solin, Chapter 42) marking the language as spoken. Among other things, this requires the use of colloquial lexical expressions (see the papers by Salmon 1987 [1967], Replogle 1987 [1973]) and also sentence structures (see Salmon (1987 [1965]). As social dialects, the language of thieves and vagabonds (“cant”) and sexual slang (“bawdy”) play a particular role (see Scheler 1982: 108–111 with many examples, the dictionary by Partridge 1955 and Musgrove 1987 [1981]). In the 19th century, Shakespeare’s linguistic permissiveness led to the publication of a purified Family Shakespeare, published by Thomas Bowdler (1807), in which “those words and expressions are omitted which cannot with propriety be read in the family”. The influence of regional dialects is to be found in Shakespeare’s plays in vocabulary and pronunciation, and to a much less extent in syntax and semantics. The features of the conventional stage dialect that he used are today found chiefly in the South-West, but they were originally found in most of the dialects south of the Thames. They included such characteristics as the voicing of initial [f] and [s], the change of initial [θr] to [dr], and the use of ich for the pronoun I, with the consequent use of the contracted forms cham ‘I am’ and chill ‘I will’ (Brook 1976: 177).
The best-known examples of a stage representation of regional dialect are the exchanges between Edgar and Oswald in King Lear, (3): (3)
Edgar: Chill not let go, zir, without vurther [cagion]. Oswald: Let go, slave, or thou di’st! Edgar: Good gentleman, go your gait, and let poor voke pass. And chud ha’ bin zwagger’d out of my life, ’twould not ha’ bin zo long as ’tis by a vortnight. Nay, come not near th’ old man; keep out, che vor’ ye, or Ice try whither your costard or my ballow be the harder. Chill be plain with you.
815
816
V Early Modern English Oswald: Out dunghill! [They fight.] Edgar: Chill pick your teeth, zir. Come, no matter vor your foins. (King Lear IV.vi.235–245) Apart from borrowing, word-formation proved to be a very important means in EModE of enriching the vocabulary. Thus, Barber in his analysis of “a 2 per cent sample of the OED” consisting of “1,848 words of reasonably certain etymology” (1997: 219; 220) comes to the conclusion that while 625 words (33.8%) are borrowings, 1,223 words (66.2%) can be attributed to word-formation processes. Out of these processes, affixation, and in particular suffixation with 607 words is the most productive process, followed at some distance by compounding with 217 words (Barber 1997: 221). By and large, the same holds true for Shakespeare, even though it looks as if in the OED more compounds are ascribed to Shakespeare than affixes. Scheler (1982: 115–128) reports in detail on a number of older German studies from the 1950s which sought to quantify Shakespeare’s compounds, prefix- and suffix-formations and conversions. All of these studies are of course limited by the reliability of the OED in correctly attributing neologisms to Shakespeare. The article by Salmon illustrates “some functions of Shakespearian word-formation” by discussing a number of examples in order “to illustrate one aspect of Shakespeare’s craftsmanship as a poet and dramatist – and assist in a more precise characterization of his style” (Salmon 1987: 194 [1970: 14]). Nevalainen (2001) provides a chapter on “Shakespeare’s new words” and deals with compounding, conversion, and affixation.
5 Grammar Apart from the two modern grammars introduced in Section 1.1, the classic monographs by Brook (1976: Chapters 3 and 4) and Hussey (1982: Chapter 4) provide general overviews of morphology and syntax (see Cowie, Chapter 38; Seoane, Chapter 39). The books by Wikberg, Yes-No Questions and Answers in Shakespeare’s Plays (1975), Burton, Shakespeare’s Grammatical Style (1973), and Houston, Shakespearean Sentences (1988) deal with more specific syntactic aspects of Shakespeare’s language. The collection of papers by Salmon and Burness (1987) offers articles on sentence structures in colloquial Shakespearean English, on auxiliaries, on multiple negation, and on the use of -eth and -es as verb endings in the First Folio. The short overview by Adamson (2001b) studies “small words” such as his or it, the auxiliaries may and shall, deictics such as this and that, etc. With regard to grammar, the language of Shakespeare shows no marked differences to that of his contemporaries. Scheler (1982: 87) is of the opinion that there is no single construction that occurs exclusively in the Shakespeare corpus. For instance, we do not find split infinitives in Shakespeare’s works, but they cannot be found in the works of his contemporaries Spenser and Kyd either. Thus, the differences between Shakespeare and other dramatists are differences in style and degree but not in kind, so that in comparison to Spenser’s old-fashioned, formal, and learned mode of expression, Shakespeare’s use of language appears to be more modern and popular. However, when we read a play by Shakespeare, we automatically compare his language use to the standards of Modern English grammar. Against this linguistic background, two features of Shakespeare’s and EModE morpho-syntax in general stick out:
51 Early Modern English: The language of Shakespeare On the one hand, the language sounds old-fashioned and quaintly German, especially in terms of the use of inversion in questions, the use of pronouns in imperatives, and the use of modals as main verbs as shown in the following citations, (4)–(6): (4)
What makes he heere (Othello I.ii.49)
(5)
go we (Much Ado about Nothing III.i.32)
(6)
I can no more (Hamlet V.ii.312; Scheler 1982: 84–85)
On the other hand, these and other constructions have counterparts and show a good deal of variation, often within a single play. Quirk (1987: 9) shows that in imperatives “[t]he pronoun may have subject form or object form or it may be absent, and in some cases […] all three possibilities can occur with the same verb”, (7)–(9): (7)
Come thou on my side. (Richard III I.iv.263)
(8)
Come thee on. (Antony and Cleopatra IV.vii.16)
(9)
Come on my right hand. (Julius Caesar I.ii.213).
Quirk (1987: 9) concludes: “While it would be idle to pretend that these three forms of imperative were always carefully distinguished in meaning at this time, we must not assume that they were usually synonymous”. Variation within a single play or even a scene can be highlighted by the use of do. Barber (1997: 193–194) illustrates the use of do in negative, interrogative, and affirmative sentences by resorting to the scene of Clarence’s murder in Richard III. Among others, he cites the following examples. In (10) and (11) we find questions formed with do, but (12) and (13) illustrate questions in which the subject is placed after the lexical verb. Similarly, negative sentences can be formed with do, as in (14), or without, as in (15) and (16): (10) How do’st thou feele thy selfe now? (Richard III I.iv.120) (11) Wherefore do you come? (I.iv.171) (12) Why lookes your Grace so heauily to day? (I.iv.1) (13) how cam’st thou hither? (I.iv.85) (14) O do not slander him. (I.iv.241) (15) He sends you not to murther me for this (I.iv.213) (16) beleeue him not. (I.iv.147)
817
818
V Early Modern English With regard to affirmative declarative sentences in this scene, Barber concludes that these “are mostly formed without do. […] There is however a substantial minority of cases where do is inserted. […] In these cases, it is highly improbable that do is inserted to give emphasis: Clarence is not saying ‘Your eyes DO menace me’. His sentence is just an alternative way of saying ‘Your eyes menace me’, and the do is probably quite unstressed” (Barber 1997: 194). In the case of do, particular findings for a Shakespearean play or the entire corpus (see Stein 1990: Chapter 7) can now be linked to the work on the development of do in EModE at large, which is outlined by Nevalainen (2006: 108) as follows: “During the Early Modern period, do first spreads to negative questions, then to affirmative questions and most negative statements as well as, to a certain extent, to affirmative statements” (see Warner, Chapter 47). In this case, another long-standing complaint that much linguistic work on Shakespeare is isolated and fragmented is fortunately no longer true. Methodologically, the treatment of functional variants needs careful consideration, because the Shakespeare corpus also shows differences over time. Therefore, “we have to bear in mind that Shakespeare’s plays were written over a period of at least twenty years in a time of extremely rapid language change, so that we cannot always assume that all the instances of a particular feature were used according to identical principles” (Grannis 1990: 106). The morphological variation between the verb endings -(e)s and -(e)th of the thirdperson singular of the present tense is a good case in point to show such changes over time in the Shakespeare corpus. When we take a look at language histories or introductory textbooks to EModE, the general development is well known and can be summarized in just one sentence. “Of northern origin, -e(s) had largely replaced the southern -(e)th in the General dialect by the seventeenth century, although -(e)th prevailed in some regional dialects and formal genres much longer” (Nevalainen 2006: 90). However, in a number of detailed works on the Shakespeare corpus, Stein has found out that the corpus shows a systematic internal divide at around 1600 because “[b]etween 1590 and 1610 the morphology of the Shakespeare corpus changes in such a systematic way that the effect of textual history in creating these patterns can be ruled out” (Stein 1987: 413–414). As a third example showing differences between PDE and EModE, modal auxiliaries were mentioned at the beginning of this section. Unlike Modern English, an EModE auxiliary could be followed by a past participle, as in the following example, (17): (17) her death was doubtfull, And but that great commaund oreswayes the order, She should in ground vnsanctified been lodg’d (Hamlet V.i.227–229; Barber 1997: 197) Barber (1997: 197) points out that the text “is taken from the good Quarto of 1604: the First Folio of 1623 reads ‘She should in ground vnsanctified haue lodg’d’ ”. He further mentions that this construction was not frequent, but that it did occur often enough not to be mistaken as a mistake or misprint. This example shows that the evaluation or interpretation of language forms is laden with difficulties that can be attributed to the authenticity of the text and the different publication histories of individual plays. Hamlet exists in two Quarto editions (Q1 from 1603 and Q2 from 1604) and in the Folio version from 1623. Q1 is not regarded as a reliable text.
51 Early Modern English: The language of Shakespeare Another syntactic peculiarity is the possibility that the lexical verb, normally following a modal auxiliary, could “be dispensed with in sentences indicating motion” (Barber 1997: 197) as in (18)
I must to Couentree (Richard II I.ii.56)
While these peculiarities may just seem odd from the point of view of PDE, they do not pose major stumbling blocks for the understanding of an utterance. However, with regard to this, the modals are situated at the crossroads between syntax, semantics, and pragmatics. Nakayasu (2009: Chapter 2) provides an overview of these different strands of earlier research on modals in EModE. For instance, Kakietek (1972) examines their use in ten Shakespeare plays by means of componential analysis, pointing out the following: “By paying attention to the context in which the modals happen to appear, the grammarian manages to avoid the common mistake of confusing the semantics of the modals with what pertains strictly to the context” (Kakietek 1972: 17). The comprehensive study by Nakayasu (2009: 243) analyzes the modals shall/should, will/ would and their contracted forms for Antony and Cleopatra, Julius Caesar, and Love’s Labors Lost by subjecting them first to “a variationist analysis of syntactic, semantic, and pragmatic aspects” and then to micro- and macro-pragmatic analyses in terms of the relationships that hold between modals and speech acts, between modals and politeness, and among dialogue, discourse, and modals.
6 Pragmatics and discourse Historical pragmatic investigations of Shakespeare’s language are more recent (for a more general view of EModE pragmatics, see Archer, Chapter 41). Excepting early studies such as Fish (1976) and Porter (1979), the approach (see Rudanko 1993) came into being in the 1990s. Blake’s (2002) grammar has a chapter on “Discourse and Register” (Blake 2002: 271–303) and on “Pragmatics” (Blake 2002: 304–325). Pragmatic features studied in Shakespeare’s language mainly cover the investigation of forms of address, politeness strategies, discourse markers, or speech acts. For a more detailed overview Busse and Busse (2010). Pragmatic studies of Shakespeare’s language cannot clearly be separated into pragmaphilological studies and those that use a function-to-form or a form-to-function mapping of diachronic pragmatics (see Jacobs and Jucker 1995). Often these approaches are combined. Some of the classic studies have also focussed on linguistic features with discourse function in context. The collection of papers in Salmon and Burness (1987) contains a number of classic studies on forms of address, e.g. Barber (1987 [1981]) and Mulholland (1987 [1967]), and on ritual utterances, such as greeting or parting formulae, e.g. Salmon (1987 [1967]). Generally speaking, the point of the social and pragmatic relevance of address behavior in Shakespeare’s plays has been frequently researched and commented on, and within this framework many studies have concentrated on pronouns since the 19th century until the present. In particular, one focus is on the use of the distinction between you and thou forms, or what is often called a T (tu) and V (vous) distinction (see Busse, Chapter 46, Section 2.5; for critical reports see U. Busse 2002: Chapter 2; Mazzon 2003; Stein 2003; Freedman 2007). The investigation of nominal forms of
819
820
V Early Modern English address or vocative constructions has been influenced by established models for explaining pronoun usage, although only rarely has the interplay between nominal and pronominal forms of address been explored comprehensively as one interlinked phenomenon (Barber 1987 [1981]; U. Busse 2002: Chapter 6; B. Busse 2006a). B. Busse (2006a) investigates the functions, meanings, and variety of forms of address in a corpus of seventeen Shakespeare plays, which are selected according to editorial, thematic, generic, synchronic, and diachronic considerations. Contrary to Brown and Gilman’s (1960) and Brown and Ford’s (1961) parameters of power and solidarity and the rigid social structure allegedly existent in Shakespeare’s time (Breuer 1983; Stoll 1989), B. Busse (2006a) claims that nominal forms of address in Shakespeare are experiential, interpersonal, and textual markers which reflect and create relationships, identity, and attitude as well as messages and habitus. They also structure the discourse, and are meaning-making within a performative context. Drawing on a new categorization of vocative forms, which includes labels like “conventional terms” or “terms referring to natural phenomena”, B. Busse (2006a) establishes the functional potential of vocative forms also from a quantitative perspective and according to synchronic, diachronic, and generic parameters and with regard to the use of vocative forms by the characters of the plays. Politeness studies of Shakespeare’s language are most frequently based on Brown and Levinson’s (1987 [1978]) theory. Blake (2002: 320–325) is a non-technical introduction to their model. He uses several pertinent examples to illustrate how politeness works. Brown and Gilman (1989) investigate four Shakespearean tragedies (Hamlet, King Lear, Macbeth, and Othello) and apply Brown and Levinson’s (1987 [1978]) three variables of power, distance and ranked extremity to discourse between two characters, which only differs in one of the three variables. Kopytko (1993, 1995) applies Brown and Levinson’s framework but explicitly excludes forms of address. He finds that in his sample of Shakespearean comedies and tragedies positive politeness strategies prevail over strategies of negative politeness. Rudanko (1993) uses Brown and Levinson’s (1987 [1978]) framework as a base but stresses the need to enlarge their framework for the historical analysis of Shakespeare’s Timon of Athens. Because “acting in a way other than politely [is not] necessarily the same as the absence of politeness” (Rudanko 1993: 167). He also develops an inventory of substrategies for nastiness (Rudanko 1993: 168–171). Culpeper (1996, 1998) applies his theory of impoliteness to Shakespeare’s dramatic dialogue in the banquet scene of Macbeth. He highlights the functional potential of impoliteness, which goes beyond entertaining the audience. Impoliteness is seen as a “symptom or as a cause of social disharmony” (Culpeper 1998: 86) and as a means of characterization. Magnusson’s (1999) historically situated comparison of politeness in Early Modern letter writing and Shakespeare’s plays uses an interdisciplinary approach that emphasizes the social exchange in Shakespeare’s language. Discourse markers are less frequently investigated for Shakespeare’s English than they are within historical pragmatics in general. Salmon (1987 [1967]) investigates colloquial English in the Falstaff plays. Blake (2002) studies interjections like o/oh, alas, ’sblood/zblood (‘God’s blood’), and why, interrogative pronouns like why, what, and well, shorter clauses or phrases like I say, for upon my life, and forms which derive from imperatives like the doubling of come. Discourse markers stress an emotional overtone and the speaker’s attitude and serve as a means of characterization. He also draws our attention to the complex interplay between editorial considerations and
51 Early Modern English: The language of Shakespeare pragmatic interpretation. For example, F1 and quarto editions of respective plays are likely to be deviating from one another due to editorial decision-making. Variants occur and the syntactic as well as pragmatic function of words like well can be interpreted differently. U. Busse’s (2002: Chapter 7) case study of prithee and pray you analyzes the functions of these discourse markers from both a quantitative and a qualitative point of view. Pray you is identified to be most frequently used as a discourse marker and most frequently collocating with the form of address sir. Also, he stresses a complex interplay between the communicative functions of prithee and pray you, which no longer sees (I) pray you as deferential and prithee as an in-group identity marker (Brown and Gilman 1989: 183–184). Although it can be assumed that what Romaine (1982: 122) called “the uniformitarian principle” is also valid for the realization of speech acts in the past history of English, the analyses of speech acts in Shakespeare have revealed a number of methodological challenges, which also generally occur when older stages of the English language are analyzed. As early as 1993, in his study of speech acts in Coriolanus, Rudanko (1993) drew our attention to the fact that two approaches need to be distinguished when studying speech acts in Shakespeare: the lexical properties of illocutionary verbs and their classification, on the one hand, and the “direction of fit” between words and the world, on the other. From a historical-pragmatic point of view, it can be said that a form-to-function mapping might not include all relevant syntactic realizations of speech acts, even if a comprehensive electronic text corpus is used. Then there is also the challenge of assigning the right communicative function to the respective form. A function-to-form mapping partly cannot immediately assume that, for example, a directive speech act is only realized by one or two syntactic forms. On the basis of an analysis of King Lear, U. Busse (2008) explains the relationships which hold between grammatical sentence types such as imperatives, on the one hand, and their communicative function on the other. In addition, as modern linguists we can only reconstruct the past and we have to be aware that the choice of available forms also depends on the period of observation. Furthermore, speech acts are fuzzy and it seems appropriate to introduce the concept of a “pragmatic space” (Jucker and Taavitsainen 2000: 73, 84) to explain their functions in context. Nonetheless, Jucker and Taavitsainen (2000) and others have illustrated that diachronic speech analysis is possible and that different realizations may in part also reflect differences in function.
7 Recent trends and further directions Recent trends in studies of Shakespeare’s language show the same dominance of corpus linguistics that is also generally visible in (English) historical linguistics. Some are highly interesting contributions to Shakespeare’s language and to the field of “modern historical linguistics” in general. These focus, for example, on key words, which are retrieved by complex statistical procedures (Culpeper 2002), and relate these characterizations to the fact that Shakespeare’s language is constructed dialogue. Others use a historical corpus linguistic approach as a method and combine it with historical-pragmatic, historical sociolinguistic, stylistic, and literary critical considerations (B. Busse 2006a, b). Those studies that focus on EModE features of the English language or take on a diachronic perspective (including a function-to-form or a form-to-function mapping) also
821
822
V Early Modern English frequently draw on Shakespeare’s language, because it is a closed set of data. For example B. Busse (2010) uses the Shakespeare corpus to discover the syntactic realizations of stance adverbials in Shakespeare and to establish their functional import. The informed interplay between linguistics and literary criticism can be seen in Munkelt and B. Busse (2007) and Biewer (2009). Recently there has also been an interest in using a multimodal approach between text and performance (B. Busse 2006a) and film adaptations of Shakespeare’s plays (McInytre 2008). Generally speaking, it can be said that some areas of Shakespeare’s language are still rather unexplored. For example, a syntactic investigation of Shakespeare’s language using modern corpus-based approaches and methodology and leading to a comprehensive grammar would be desirable. Also, pragmatic features, such as speech acts, are less well charted, or still unexplored, as, for instance, the investigation of discourse markers or modality in the entire Shakespeare corpus. Those features will hopefully be studied in the Shakespeare corpus by including socio-historical and also cognitive aspects and by relating them to pragmatic considerations.
8 References Abbott, Edwin A. 1966 [31870]. A Shakespearian Grammar: An Attempt to Illustrate Some of the Differences Between Elizabethan and Modern English. London: Macmillan [Repr. New York: Dover, 1966] Available at http://test.perseus.tufts.edu/hopper/text.jsp?doc=Perseus:text:1999. 03.0080 Adamson, Sylvia. 2001a. The grand style. In: Adamson et al. (eds.), 31–50. Adamson, Sylvia. 2001b. Understanding Shakespeare’s grammar: Studies in small words. In: Adamson et al. (eds.), 210–236. Adamson, Sylvia, Lynette Hunter, Lynne Magnusson, Ann Thompson, and Katie Wales (eds.). 2001. Reading Shakespeare’s Dramatic Language: A Guide. London: The Arden Shakespeare [Cengage Learning]. Barber, Charles. 1987 [1981]. You and thou in Shakespeare’s Richard III. In: Salmon and Burness (eds.), 163–179. (First published in Leeds Studies in English, New Series 12 [1981]: 273–289.) Barber, Charles. 1997. Early Modern English. 2nd edn. Edinburgh: Edinburgh University Press. Bartlett, John. 1894. A New and Complete Concordance or Verbal Index to Words, Phrases, and Passages in the Dramatic Works of Shakespeare. London: Macmillan. Biewer, Carolin. 2009. Dietetics as a key to language and character in Shakespeare’s comedy. English Studies: A Journal of English Language and Literature 90(1): 17–33. Blake, Norman F. 1983. Shakespeare’s Language: An Introduction. Basingstoke: Macmillan. Blake, Norman F. 1990. Shakespeare’s language: Some recent studies and future directions. Deutsche Shakespeare-Gesellschaft West. Jahrbuch 1990, 61–77. Blake, Norman F. 2002. A Grammar of Shakespeare’s Language. Houndmills: Palgrave. Bowdler, Thomas. 1807. The Family Shakespeare. London: Longman. Bradbrook, Muriel C. 1954. Fifty years of the criticism of Shakespeare’s style: A retrospect. Shakespeare Survey 7: 1–11. Breuer, Horst. 1983. Titel und Anreden bei Shakespeare und in der Shakespearezeit. Anglia 101: 49–77. Brook, George L. 1976. The Language of Shakespeare. London: Deutsch. Brown, Penelope and Stephen C. Levinson. 1987 [1978]. Politeness: Some Universals in Language Usage. Cambridge: Cambridge University Press. Brown, Roger W. and Albert Gilman. 1960. The pronouns of power and solidarity. In: Thomas A. Sebeok (ed.), Style in Language. Cambridge, MA: MIT Press.
51 Early Modern English: The language of Shakespeare Brown, Roger W. and Albert Gilman. 1989. Politeness theory and Shakespeare’s four major tragedies. Language in Society 18: 159–212. Brown, Roger W. and Marguerite Ford. 1961. Address in American English. Journal of Abnormal and Social Psychology 62: 375–385. Burton, Dolores M. 1973. Shakespeare’s Grammatical Style: A Computer-assisted Analysis of Richard II and Antony and Cleopatra. Austin: University of Texas Press. Busse, Beatrix. 2006a. Vocative Constructions in the Language of Shakespeare. Amsterdam/Philadelphia: John Benjamins. Busse, Beatrix. 2006b. Linguistic aspects of sensuality: A corpus-based approach to will-construing contexts in Shakespeare’s works. In: Christoph Houswitschka, Gabriele Knappe, and Anja Mu¨ller (eds.), Anglistentag 2005 Bamberg: Proceedings, 125–143. Trier: WVT. Busse, Beatrix. 2010. Adverbial expressions of stance in Early Modern “Spoken English”. In: Jo¨rg Helbig (ed.), Anglistentag 2009 Klagenfurt: Proceedings, 47–64. Trier: WVT . Busse, Ulrich. 2002. Linguistic Variation in the Shakespeare Corpus: Morpho-syntactic Variability of Second Person Pronouns. Amsterdam/Philadelphia: John Benjamins. Busse, Ulrich. 2007. Review of Blake (2002) and Hope (2003). Journal of Historical Pragmatics 8(1): 127–137. Busse, Ulrich. 2008. An inventory of directives in Shakespeare’s King Lear. In: Andreas H. Jucker and Irma Taavitsainen (eds.), Speech Acts in the History of English, 85–114. Amsterdam/ Philadelphia: John Benjamins. Busse, Ulrich and Beatrix Busse. 2010. Shakespeare. In: Andreas H. Jucker and Irma Taavitsainen (eds.), Historical Pragmatics, 247–281. (Handbooks of Pragmatics, 8.) Berlin/New York: Mouton de Gruyter, Cercignani, Fausto. 1981. Shakespeare’s Works and Elizabethan Pronunciation. Oxford: Clarendon Press. Crystal, David. 1988. The English Language. Harmondsworth: Penguin. Crystal, David. 2003. The language of Shakespeare. In: Stanley Wells and Linda Cowen Orlin (eds.), Shakespeare: An Oxford Guide, 67–78. Oxford: Oxford University Press. Crystal, David. 2005. Pronouncing Shakespeare: The Globe Experiment. Cambridge: Cambridge University Press. See: The Shakespeare Portal. http://www.theshakespeareportal.com Crystal, David. 2008. Think on My Words: Exploring Shakespeare’s Language. Cambridge: Cambridge University Press. Crystal, David and Ben Crystal. 2002. Shakespeare’s Words: A Glossary and Language Companion. Harmondsworth: Penguin. Culpeper, Jonathan. 1996. Towards an anatomy of impoliteness. Journal of Pragmatics 25: 349–367. Culpeper, Jonathan. 1998. (Im)politeness in dramatic dialogue. In: Jonathan Culpeper, Mick Short, and Peter Verdonk (eds.), Exploring the Language of Drama: From Text to Context, 83–95. London: Routledge. Culpeper, Jonathan. 2002. Computers, language and characterisation: An analysis of six characters in Romeo and Juliet. In: Ulla Melander-Marttala, Carin Ostman, and Merja Kyto¨ (eds.), Conversation in Life and in Literature: Papers from the ASLA Symposium, 11–30. Uppsala: Association Sue´doise de Linguistique Applique´e. Cunliffe, Richard John. 1910. A New Shakespearean Dictionary. London: Blackie. Dobson, E. J. 1968. English Pronunciation 1500–1700. 2 vols. Oxford: Clarendon Press. Evans, Ifor. 1952. The Language of Shakespeare’s Plays. London: Methuen. Evans, G. Blakemore with the assistance of J. J. M. Tobin (ed.). 1997 [1972]. The Riverside Shakespeare. 2nd edn. Boston: Houghton Mifflin. Finkenstaedt, Thomas, Erst Leisi, and Dieter Wolff (eds.). 1970. A Chronological English Dictionary. Heidelberg: Winter. Fish, Stanley E. 1976. How to do things with Austin and Searle: Speech act theory and literary criticism. Modern Language Notes 91: 983–1025.
823
824
V Early Modern English Fitzmaurice, Susan M. and Irma Taavitsainen (eds.). 2007. Methods in Historical Pragmatics. Berlin/New York: Mouton de Gruyter. Folkerth, Wes. 2002. The Sound of Shakespeare. London: Routledge. Franken, Gereon. 1995. Systematische Etymologie: Untersuchung einer ‘Mischsprache’ am Beispiel des Shakespeare-Wortschatzes. Heidelberg: Winter. Franz, Wilhelm. 1986 [1898/99, 1939]. Die Sprache Shakespeares in Vers und Prosa unter Beru¨cksichtigung des Amerikanischen entwicklungsgeschichtlich dargestellt: Shakespeare Grammatik. 4th edn. Tu¨bingen: Niemeyer. Freedman, Penelope. 2007. Power and Passion in Shakespeare’s Pronouns: Interrogating ‘you’ and ‘thou’. Aldershot: Ashgate. Garner, Bryan A. 1987a [1982]. Shakespeare’s Latinate neologisms. In: Salmon and Burness (eds.), 207–228. (First published in Shakespeare Studies 15 [1982]: 149–170.) Garner, Bryan A. 1987b [1983]. Latin-Saxon hybrids in Shakespeare and the Bible. In: Salmon and Burness (eds.), 229–234. (First published in Studies in the Humanities 10 [1983]: 39–44.) Go¨rlach, Manfred. 1994. Einfu¨hrung ins Fru¨hneuenglische. 2nd edn. Heidelberg: Winter. Grannis, Oliver. 1990. The social relevance of grammatical choice in Shakespeare. Deutsche Shakespeare-Gesellschaft West, Jahrbuch 1990, 105–118. Hope, Jonathan. 2003. Shakespeare’s Grammar. London. The Arden Shakespeare [Thomson Learning]. Houston, John Porter. 1988. Shakespearean Sentences: A Study in Style and Syntax. Baton Rouge: Louisiana State University Press. Hulme, Hilda M. 1962. Explorations in Shakespeare’s Language: Some Problems of Word Meaning in the Dramatic Text. London: Longman. Hulme, Hilda M. 1972. Yours that Read Him: An Introduction to Shakespeare’s Language. London: Ginn. Hussey, Stanley S. 1982. The Literary Language of Shakespeare. London: Longman. Jacobs, Andreas and Andreas H. Jucker. 1995. The historical perspective in pragmatics. In: Jucker (ed.), 3–33. Jucker, Andreas H. (ed.). 1995. Historical Pragmatics: Pragmatic Developments in the History of English. Amsterdam/Philadelphia: John Benjamins. Jucker, Andreas H. and Irma Taavitsainen. 2000. Diachronic speech act analysis: insults from flyting to flaming. Journal of Historical Pragmatics 1(1): 67–95. Kakietek, Piotr. 1972. Modal Verbs in Shakespeare’s English. Poznan´: Adam Mickiewicz University Press. Kellner, Leon. 1922. Shakespeare-Wo¨rterbuch. Leipzig: Tauchnitz. Kermode, Frank. 2000. Shakespeare’s Language. London: Allen Lane. Ko¨keritz, Helge. 1953. Shakespeare’s Pronunciation. New Haven, CT: Yale University Press. Kopytko, Roman. 1993. Polite Discourse in Shakespeare’s English. Poznan´: Adam Mickiewicz University Press. Kopytko, Roman. 1995. Linguistic politeness strategies in Shakespeare’s plays. In: Jucker (ed.), 515–540. Lass, Roger. 1999. Phonology and morphology. In: Roger Lass (ed.) The Cambridge History of the English Language. Vol. III. 1476–1776, 56–186. Cambridge: Cambridge University Press. Lass, Roger. 2001. Shakespeare’s sounds. In: Adamson et al. (eds.), 256–268. Leisi, Ernst. 1997. Problemwo¨rter und Problemstellen in Shakespeares Dramen. Tu¨bingen: Stauffenburg. Magnusson, Lynne. 1999. Shakespeare and Social Dialogue: Dramatic Language and Elizabethan Letters. Cambridge: Cambridge University Press. Magnusson, Lynne. 2007. A pragmatics for interpreting Shakespeare’s sonnets 1 to 20: Dialogue scripts and Erasmian intertexts. In: Fitzmaurice and Taavitsainen (eds.), 167–184. Mair, Christian. 2006. Twentieth-century English: History, Variation, and Standardization. Cambridge: Cambridge University Press.
51 Early Modern English: The language of Shakespeare Mazzon, Gabriella. 2003. Pronouns and nominal address in Shakespearean English: A socio-affective marking system in transition. In: Taavitsainen and Jucker (eds.), 223–249. McDonald, Russ. 2001. Shakespeare and the Arts of Language. Oxford: Oxford University Press. McIntyre, Dan. 2008. Integrating multimodal analysis and the stylistics of drama: A multimodal perspective on Ian McKellen’s Richard III. Language and Literature 17(4): 309–334. Muir, Kenneth and S. Schoenbaum (eds.). 1971. A New Companion to Shakespeare Studies. London: Cambridge University Press. Mulholland, Joan. 1987 [1967]. Thou and you in Shakespeare: A study in the second person pronoun. In: Salmon and Burness (eds.), 153–161. (First published in English Studies 48 [1967]: 34–43.) Munkelt, Marga and Beatrix Busse. 2007. Aspects of governance in Shakespeare’s Edward III: The quest for personal and political identity. In: Sonja Fielitz (ed.), Literature as History. History as Literature, 103–121. Frankfurt am Main: Peter Lang. Musgrove, S. 1987 [1981]. Thieves’ cant in King Lear. In: Salmon and Burness (eds.), 245–253. (First published in English Studies 62 [1981]: 5–13.) Nakayasu, Minako. 2009. The Pragmatics of Modals in Shakespeare. Frankfurt am Main: Peter Lang. Neuhaus, H. Joachim forthc. Shakespeare Database. http://www.shkspr.uni-muenster.de/ Nevalainen, Terttu. 2001. Shakespeare’s new words. In: Adamson et al. (eds.), 237–255. Nevalainen, Terttu. 2006. An Introduction to Early Modern English. Edinburgh: Edinburgh University Press. Onions, Charles T. 1911. A Shakespeare Glossary. 3rd edn., enlarged and revised throughout by Robert D. Eagleston. Oxford: Clarendon Press. Parod, William. 2004–10. WordHoard. Northwestern University and Nu-IT Academic Technologies. http://wordhoard.northwestern.edu/userman/whatiswordhoard.html Partridge, Astley Cooper. 1955. Shakespeare’s Bawdy. London: Routledge. Partridge, Astley Cooper. 1979. Shakespeare’s English: A bibliographical survey. Poetica 11: 46–79. Porter, Joseph A. 1979. The Drama of Speech Acts: Shakespeare’s Lancastrian Trilogy. Berkeley: University of California Press. Quirk, Randoph. 1987 [1971, 1974]. Shakespeare and the English Language. In: Salmon and Burness (eds.), 1987: 3–21. (First published in Kenneth Muir and S. Schoenbaum (eds.), A New Companion to Shakespeare Studies, 67–82. London: Cambridge University Press, 1971. Repr. with minor alterations in Randoph Quirk, The Linguist and the English Language, 46–64. London: Arnold, 1974.) Replogle, Carol A. H. 1987 [1973]. Shakespeare’s salutations: A study in stylistic etiquette. In: Salmon and Burness (eds.), 101–115. (First published in Studies in Philology 70 [1973]: 172–186.) Romaine, Suzanne. 1982. Socio-historical Linguistics: Its Status and Methodology. Cambridge: Cambridge University Press. Rubinstein, Frankie. 1984. A Dictionary of Shakespeare’s Sexual Puns and Their Significance. London: Macmillan. Rudanko, Juhani M. 1993. Pragmatic Approaches to Shakespeare: Essays on Othello, Coriolanus and Timon of Athens. Lanham: University Press of America. Salmon, Vivian. 1987 [1965]. Sentence structures in colloquial Shakespearian English. In: Salmon and Burness (eds.), 265–300. (First published in Transactions of the Philological Society 1965: 105–140.) Salmon, Vivian. 1987 [1967]. Elizabethan colloquial English in the Falstaff Plays. In: Salmon and Burness (eds.) 37–70. (First published in Leeds Studies in English 1 [1967]: 37–70.) Salmon, Vivian. 1987 [1970]. Some functions of Shakespearian word-formation. In: Salmon and Burness (eds.), 193–206. (First published in Shakespeare Survey 23 [1970]: 13–26.) Salmon, Vivian. 1987. Introduction. In: Salmon and Burness (eds.), xii–xxii. Salmon, Vivian and Edwina Burness (eds.). 1987. A Reader in the Language of Shakespearean Drama. Amsterdam/Philadelphia: John Benjamins.
825
826
V Early Modern English Scha¨fer, Ju¨rgen. 1973. Shakespeares Stil: Germanisches und romanisches Vokabular. Frankfurt am Main: Athena¨um. Scha¨fer, Ju¨rgen. 1980. Documentation in the O.E.D. Shakespeare and Nashe as Test Cases. Oxford: Clarendon Press. Scheler, Manfred. 1982. Shakespeares Englisch: Eine sprachwissenschaftliche Einfu¨hrung. Berlin: Schmidt. Schlauch, Margaret. 1987 [1965]. The social background of Shakespeare’s malapropisms. In: Salmon and Burness (eds.), 71–99. (First published in Poland’s Homage to Shakespeare 1965, 203–231.) Schmidt, Alexander and Gregor Sarrazin. 1971 [1874/1875]. Shakespeare-Lexicon: A Complete Dictionary of all the English Words, Phrases and Constructions in the Works of the Poet. 2 vols., 6th unchanged edn. Berlin/New York: de Gruyter. Smith, Bruce R. 1999. The Acoustic World of Early Modern England. Chicago: Chicago University Press. Sokol, B. J. and Mary Sokol. 2000. Shakespeare’s Legal Language: A Dictionary. London: Athlone Press. Spevack, Marvin. 1968–80. A Complete and Systematic Concordance to the Works of Shakespeare. 9 vols. Hildesheim: Olms. Spevack, Marvin. 1972. Shakespeare’s English: The core vocabulary. Review of National Literatures 3: 106–122. Spevack, Marvin. 1993. A Shakespeare Thesaurus. Hildesheim: Olms. Spevack, Marvin. 2002. [email protected]. The Shakespeare Newsletter 254 (Fall 2002) (1): 82–86. Stein, Dieter. 1987. At the crossroads of philology, linguistics and semiotics: Notes on the replacement of th by s in the third person singular in English. English Studies 68: 406–432. Stein, Dieter. 1990. The Semantics of Syntactic Change: Aspects of the Evolution of do in English. Berlin/New York: Mouton de Gruyter. Stein, Dieter. 2003. Pronominal usage in Shakespeare: Between sociolinguistics and conversational analysis. In: Taavitsainen and Jucker (eds.), 251–307. Stoll, Rita. 1989. Die nicht-pronominale Anrede bei Shakespeare. Frankfurt am Main: Peter Lang. Taavitsainen, Irma and Susan Fitzmaurice. 2007. Historical pragmatics: What it is and how to do it. In: Fitzmaurice and Taavitsainen (eds.), 11–36. Taavitsainen, Irma and Andreas H. Jucker (eds.). 2003. Diachronic Perspectives on Address Term Systems. Amsterdam/Philadelphia: John Benjamins. Wikberg, Kay. 1975. Yes-No Questions and Answers in Shakespeare’s Plays: A Study in Text Lin˚ bo: A ˚ bo Academi. guistics. A Williams, Gordon. 1997. A Glossary of Shakespeare’s Sexual Language. London: Athlone Press.
Ulrich Busse, Halle/Saale (Germany) and Beatrix Busse, Heidelberg (Germany)
VI Late Modern English 52 Late Modern English: Phonology 1 2 3 4
Introduction Vowel system developments Consonantal developments References
Abstract Using the large data set provided by contemporary observers, this paper sets out to provide a brief description of some of the developments in the segmental phonology of English which occur between 1700 and 1900. Among the major vowel developments we see a BIT/BEAT and FOOT/STRUT demerger, alongside a BATH/TRAP split, while there is also evidence for a merger involving MAT/MET words. Later in the period we see new diphthongs developed in SAY and GO words especially in Metropolitan usage. Among consonantal developments we find post vocalic [r] suppression and unetymological insertion, the almost canonical fricativization of voiced and voiceless obstruents in -ion environments, the insertion of [j] and [w] glides word initially after obstruents, as well as much by the way of syllable-initial [h]-loss and unetymological insertion.
1 Introduction The student of LModE phonology is particularly fortunate in the nature and extent of the data available both for any tentative reconstruction of contemporary pronunciation and for an assessment of ongoing phonological change in the period. Sources range from extensive pronouncing dictionaries, grammar books, phonetic commentaries, all often supported and enhanced by innovative and revealing orthographic innovations (Jones 2003: Chapter 2). While some of the source materials are seen by some as devalued because of their heavy prescriptive content (despite the evidence this provides for ongoing sociophonetic processes), this can be over-emphasized and there are as many commentaries which are based on sophisticated and insightful phonetic and phonological modeling (Jones 1991: 11–14).
2 Vowel system developments 2.1 High front vowels and the
BIT/BEAT
split
For the majority of 18th century observers the “English Vowel Shift” (see Krug, Chapter 48) raising of [ee] to [ii] has been completed, creating meat/meet and beat/beet mergers. Only Tiffin (1751) seems unique in the early 18th century in keeping distinct high front vowel pronunciations in items like beet and beat, which he seems to distinguish in terms of vowel height. It is difficult to give a precise value to these two vowels but we might speculate, in the light of the fact that he sets meet against it/meat examples, that a Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 827–842
828
VI Late Modern English contrast in relative height is involved, something like [i]/[i̞] (or perhaps [ɨ]), with bit and it showing the latter value, rather than some kind of [ɪ]. For Walker (1791: Section 246) the meet/meat difference is “very trifling”, one which is “distinguishable to a nice ear, in the different sounds of the verbs, to feel and to meet and the nouns flea and meat. This has always been my opinion; but upon consulting some good speakers on the occasion, and in particular Mr. Garrick, who could find no difference in the sounds of those words, I am less confident in giving it to the public”. Some kind of difference in pronunciation between beet/beat types is also hinted at by Entick (1795: 3): “many words are confounded in the English language, which seem to have a sufficient claim to be distinguished: such as thee, the; meet, meat; beet, beat; and many others that might be enumerated”, suggesting, perhaps, a lexical constraint upon the process. Such constraints seem to underlie the well-recorded “overshoot” raising cases of [ii] in items like break and steak which fail to conform to the dictates of the English Vowel Shift by remaining in a mid vowel state in the modern language, with some observers in the 18th century recording [brik], [ɡrit] manifestations (Wyld 1937: 172–173; Ellis 1869: 88–89). Another issue in this area of the vowel phonology arises as to whether there exists in the 18th century some kind of tense/lax [i]/[ɪ] contrast of the type found in modern Standard English between items such as beat, beet on the one hand, and bit, hit on the other. The evidence suggests that in the early 18th century there was no lax, centralized [ɪ] vowel in the phonology for hit, bit words; rather, the vowel space in such items was seen as “close” to that in beet/beat, and may have been something like [i̞], as Wyld (1953: 207) observes: “the long forms with [i] were far commoner during the first four centuries of the Modern period than at present. ‘Peety’ [piti] for pity was occasionally heard until quite recently, and ‘leetle’ [litl] is still used facietiously in the sense of ‘very little’ ”, although Dobson (1968: 570 n.) is equally adamant that Middle English [ɪ] is maintained throughout the 16th and 17th centuries. Many commentators do not appear to recognize – or hear – a clear phonetic contrast of the modern bit/beet type. Rather, the evidence they present seems to suggest that the vowels in bit and beet are “near alikes”, with the former showing no clear signs if any kind of centering whatsoever. Gildon and Brightland (1711: 17 n. 12) state quite unequivocally that “When (i) is short, it sounds most commonly like the (i) of the French and other Nations, with the small Sound”. As a whole, the evidence from the 18th century points to a raised and high value for the vowel in words like sit and fit, as commonplace, and not merely perceived as a non-standard, and certainly regional pronunciation, one particularly characteristic of the “Scotch” dialect (Jones 1995: 113–120) where we find Buchanan (1770: 45) pontificating: “I shall adduce but a few examples, out of a multitude, to shew how North-Britons destroy just quantity, by expressing the long sound for the short, and the short for the long […] as ceevil for cɪ̆ vil”. In general, 18th century commentators point to the vowel space in words like bit, hit and the like as “close to” [i], rather than to any lowered, centralized [ɪ] output. The 19th century evidence for an [i]\[ɪ] contrast is considerably stronger. Batchelor (1809: 6) uses a specific untitled graph for bit words which gives witness, at least by implication, to the emergence of a BEAT/BIT split. That this “short” i has, indeed, a distinct phonetic value is perhaps suggested by its comparison with the French [u¨]: “The sound of the French u is unknown in England, though it is said to be common in Scotland, in such words as geud (good). It is a slender sound, and possesses some alliance
52 Late Modern English: Phonology with the English i in pin, and the unaccented u […] It is said to require a considerable pouting of the lips, to pronounce it properly” (Batchelor 1809: 24), and although he does not go quite as far as to equate the two vowel sounds, he does suggest an affinity between short i and the unaccented u. Cooley’s (1861) phonotype system distinguishes long or open vowels, from stopped or derivative types, an instance of the former being the open in items like beat, eel, contrasting with the stopped in ill, pit, him types. Ellis (1874: 1280–1287) in his On Paleotype seems quite confident that such a contrast does now indeed exist: “Distinguish English finny, French fini (fini, fini)”. For Sweet (1877: 110–111), the lowering tendency of the short i is one of its defining characteristics and some of his evidence does seem to suggest that the hit vowel could be closer to [ɛ] than [e].
2.2 Happy tensing Wells (1982: 257) records the increasing tendency in speakers of both modern British and American English to give a tensed, close high [i] value to the final vowel in words like happy, lucky, and coffee, observing that “[w]here and when the [i] pronunciation arose is not certain. It has probably been in use in provincial and vulgar speech for centuries”. The phenomenon seems to be clearly (if not regularly) attested by several observers in the first half of the 18th century. Gildon and Brightland (1711: 14), in the midst of their discussion of long (e), note that, at the end of words this sound is variously expressed, by (ee) as in Pharisee, agree, and by (ea) as in sea, flea, pea. Bailey’s (1726: 56–58) Table of Words written different from their Pronunciation seems to suggest rather strongly that this “happy tensing” is at least being introduced into the phonology at this date, since he provides respellings like hunnee ‘honey’, jepurdee ‘jeopardy’, luckee ‘lucky’, mellancollee ‘melancholy’, munnee ‘money’, and munkee ‘monkey’. Likewise Owen (1732: 101–106), in a similar list, shows pronunciations like furmetee ‘frumenty’, hunnee ‘honey’, jepurdee ‘jeopardy’, lackee ‘lackey’, and many others.
2.3 The
BATH/TRAP
split
In several regional variants of Modern English there exists a qualitative and quantitative vowel value contrast between words like bath, father, rather and cat, sat, and fat. The contrast in question is often between segments such as [ɑ] and [æ], respectively. While the historical evidence is extremely difficult to interpret, it would appear that only rarely before the beginning of the 18th century can we find any evidence of the “splitting” of what was originally a shared low vowel (Beal 1999: 105–119; MacMahon 1998: 1143–1146). Even quite late in the century, commentators like Scott (1774) and Entick (1795) only show “short a” – presumably [a] – values for items such as cat, bat, alm, balm, dance, rather, fall, and even haunch. Now, if it were the case that in the London dialect there was developing some contextually conditioned contrast between not only quantitatively different low vowels, but also one which involved some kind of qualitative distinction as well, say between [a] and [ɑ]/[ɑɑ], then it is one which might be most salient for those regional speakers, like the Scots, for whom any such distinction did not exist, or was in some way different (McKnight 1928: 453). What might be a hint that at least some kind of quantitative contrast is emerging is provided by the observation of the Scot Mitford (1774: 34) that “A open is the
829
830
VI Late Modern English usual sound of short a. It is long in father, after, slander, and in dance, enhance, advance, and some other words formerly spelt with au”. Another Scot – James Douglas – long resident in London and familiar with the speech of the upper classes contrasts an A4 vowel in items like hard, glass, fast, and advance with an A2 in mad and bad (Douglas 1740). Perhaps the most complete observation is that of his Scottish namesake Sylvester Douglas (1779) who records the a sound in aunt which is less open than the a in father, which in its turn is different in both quality and quantity from the a in ant, evidence surely suggestive of a BATH/TRAP split in the Pure (Metropolitan) dialect, while it would appear that Scots speakers accommodate to something close to [ɔ] when attempting to produce the standard English form in words like aunt and father. James Douglas (much like William Johnston in his 1764 A Pronouncing and Spelling Dictionary) records, for the first time in the period, a set of alternations, phonetically conditioned, affecting the outputs of “short a”, suggesting that the alternation between the front and back a sound is a function of the consonantal segment which terminates the syllable: relative length and backness are associated in particular with terminations in fricatives and sonorants. Likewise, in the early 19th century for Batchelor (1809), the “long” version of the a sound is mainly, if not even canonically, confined to pre [r] and [r] + consonant environments, although syllables terminating in [l], [s] plus consonant, and the vowel outcome of the Middle English [au] diphthong are also candidate contexts for this vowel. The examples cited in his Vocabulary, Consisting Principally of Accidental Errors in Pronunciation (Batchelor 1809: 119–147) containing the long a symbol include narrow, arbor, argue, largess, learning, service, storm; daughter, drawn, and jaundice, while short a values appear mainly in pre-obstruent contexts: aks ‘asks’, skrat ‘scratch’, stab ‘stob’, krap ‘crop’, backard ‘backward’. Yet there is considerable evidence to suggest that the BATH/TRAP split was affected by lexical as much as phonetic triggers (Savage 1833). Not only that, but the use of the low back variant was often seen as a “vulgarism” to be avoided in the best society. Decrying the observations of “a teacher of elocution” – most probably Sheridan – who recommends pronunciations of basket, castle grant and sample by whatever is signified but the notations bas-ket, cas-tle, gr-ant, and sam-ple, Savage (1833: xxvi–xxvii) becomes quite aggravated, declaiming: “What if a Fribble should say to us, I put the plant with the glas into the bas-ket which was gr-anted me and carried it as a sam-ple to the castle. – We ask of every man if it would not disgust him to sickness”. Ellis (1874: 1148) notes that the vowel contrast was open to considerable variation, and we can almost sense his frustration at what looks more and more like an “anything goes” situation in this area of the phonology when he writes: “The words class, staff, demand, are pronounced with (aa, a, ah, aah, æ, ææ), by different careful speakers, and even (ah, Oh) are on occasion used by others” (Ellis 1869: 68 n.). The overall situation in the late 18th century appears to be much as Labov (1994: 33) suggests it is in modern British English: “The only true path for learning the broad a class is to absorb it as a set of brute facts as a first language learner, or failing that, to be enrolled in a British public school in early childhood”.
2.4 The
MAT/MET
merger
There is also extensive evidence from near-alike lists relating to short a segment to suggest that the vowel has a raised F2 quality with the concomitant possibility of a
52 Late Modern English: Phonology MAT/MET-type merger. Although many of the instances cited involve syllable final [r], where a lowering of [ɛ] to [a] might be expected, not all cases are by any means like this and the [a]/[ɛ] equivalence can be seen in many different environments. Thus, for example, Bailey’s (1726) lists contain: yarn/earn; earth/hearth, Harland (1719) shows: aliment/element; assay/essay; apologue ‘a fable’/epilogue; assart ‘to lop trees’/assert; waist/wast ‘has been’; waste/west; allective/ellective. Harland shows fallow/fellow, Hammond (1744) reddish/raddish; then/than, Watts (1721) belcony/ balcony; rack/wreck; vassal/vessel, and Owen (1732) channel/kennel; mash/mesh. Evidence like this might suggest that the early 18th century value for short a is something like [æ] or even [ɛ̞ ]. Walker (1791: 12) and Kenrick (1784: 63) record this usage as “corrupt”, and characteristic of “flirting females and affected fops”. Fogg (1796: 168) claims that Londoners adopt “new modes that incline to feebleness” when “short a is often confounded with e; man pronounced as men; fat as fet; sand as send: this is objectionable both as emasculating the language, and unfortunately obscuring its sense, making more equivocal words that any other corruption would”. Such a merger may be what the anonymous author of The Vulgarities of Speech Corrected (Anon. 1826: 20) sees as the “miserable and disgusting affectation”: “The letter ‘a’, for example, when pronounced short, has properly a sound intermediate between ‘e’ in ‘fell’, and ‘a’ in ‘fall’, but a Scotsman endeavouring to speak English almost uniformly mistakes it for the first, and pronounces, ‘bad’ – ‘bed’, ‘tax’ – ‘tex’, ‘lamb’ – ‘lemb’ […]”.
2.5 Labial vowels Contemporary observers regard vowel sounds in this area of the phonology as “difficult”. For example, very late in the 18th century we find Walker commenting on the vagaries of pronunciation in the stressed vowels of words like bull and pull which, he claims are “sufficient to puzzle Englishmen who reside at any distance from the capital, and to make the inhabitants of Scotland and Ireland […] not infrequently the jest of fools” (Walker 1791: 173). The issues we shall see which arise most clearly from their statements relate to two main areas: a. the phonetic shape of the high back labial vowel itself, whether it was durationally long or short or had more than one qualitative value (Ekwall 1975: 50–51; Wyld 1953: 234–236); b. the extent to which (if any) a FOOT/STRUT split has taken place through a lowering and centralizing of a short [u] labial segment. The bulk of the evidence from the 18th century materials seems to point to three manifestations for the labial vowel: [u]/[uu], [ə]/[ʌ] together with what might be described as a (lexically or socially determined) “intermediate”, “compromise”, or “fudged” form between these – some kind of [ɤ] segment, perhaps Walker’s middle or obtuse u (1861: xxxiii note 113). Vide some Modern English Midland pronunciations with [ʌ]like vowels in items like butcher and cushion, which should show an historical and Standard English [ʊ] vowel; but such a vowel sound is so associated in some sociogeographic contexts with low prestige speakers that [bʌtʃə] and [bɤtʃə] pronunciations are heard, perhaps especially from those groups of speakers who are most socially
831
832
VI Late Modern English conscious of the social value of language (Trudgill 1986: 60–61; Chambers and Trudgill 1980).
2.6 The
FOOT/STRUT
split
There is evidence, even from the early 18th century, that ME [u] was being lowered and centralized into some kind of [ə]/[ʌ] vowel. The short version of the vowel Jones (1701: 110) most often illustrates as occurring in unstressed syllable contexts, together with its likeness to French e Feminine, suggesting some kind of low, central schwa value in [ə]. But Jones is careful to limit this short u sound to a very small set of lexical items: but, cut, come, done, some, son and in unstressed contexts (spelt with o) like gambol, symbol, kingdom, income, fulsome, kingdom. Even in the late 17th century, The Writing Scholars Companion (Ekwall [ed.] 1911) seems to suggest the existence of some kind of [u]/[ə]\[ʌ] alternation in what look like well-defined (mainly in continuant consonant) positions: O is obscure, like (oo) or short (u): And that, (1) Before (m), as come, kingdom, pomel, company, some. (2) before (n), as London, conduit, beyond, (3) before (l) as colour, colander. (4) before (p) as bishop; but not in words of one syllable, as shop, slop (5) before (th) as Brother, Mother, smother, doth. (6) before (ve) as above, dove, love, move; so in plover, shovel, lover. (7) after (w) as woman (but sounds (i) in women), world, worship, sword, &c. (Ekwall [ed.] 1911: 28).
Although no separate description of “short u” seems to be given in the text, that it is equated with the o graph in items like Apron, Citron, Iron, etc., suggests at least a segment lower than [u] or one even lower and centralized like [ʌ]. We might, therefore, just be entitled to interpret comments like “(oo) sounds like (u) short, in food, flood, good, hood, foot, stood, wood, wool” as evidence for a nascent FOOT/STRUT split (Dobson 1968: 360; Beal 2004: 142–145), perhaps through some intermediate [ɤ] stage. Later in the century evidence for a FOOT/STRUT split is becoming clearer. The Irish Spelling Book (Anon. 1740: 61) describes the “short u” as being “form’d in the Throat, by the Larynx vibrating the Breath, and, with a moderate opening of the Lips, making a bare Murmur; as, in Nut”. A description perhaps more evocative of [ʌ] than [ʊ]. By the 19th century most observers clearly record a FOOT/STRUT split at least in metropolitan usage, and they regularly comment upon the failure of [ʊ] to lower and centralize in provincial, particularly Northern, dialects. Batchelor’s (1809: 23) comments are fairly typical: “The three kinds of u, as in but, pull, and rostrum, are seldom pronounced incorrectly; but in several parts of England, the sound of u, as in but, is exchanged improperly for the soft u in bull; as in the following instances: Come, some, pump, jump, rum”.
2.7 New diphthongs There is very little by way of evidence in the works of 18th century commentators to suggest any kind of diphthongal value for mid front and back vowels, certainly nothing corresponding to modern outputs like [feɪs] face and [ɡoʊ] go. Such a development “in the precursor of RP seems to have happened around 1800” (Wells 1982: 210), although the evidence suggests that such an early recognition of the phenomenon was sporadic
52 Late Modern English: Phonology and uncertain, and that diphthongal forms were only firmly established in prestige speech by the middle of the 19th century at the earliest. In the 1791 edition of his A Critical Pronouncing Dictionary, under a discussion of the “diphthong” ei (normally assigned a value of “long slender a” as in deign, reign, vain, etc.), Walker (1791: 40) comments that when it is found in historical pre-[c¸] contexts in items like eight, weight, “ei, followed by gh, sounds both vowels like ae; or if we could interpose the y consonant between the a t and in eight, weight, &c. it might perhaps convey the sound better. The difference, however, is so delicate as to render this distinction of no great importance”. This diphthongization may be lexically confined to a few words such as weight and eight, and Walker’s comments tend to suggest too that any [e]/[eɪ] contrast may well be below the range of observability. Yet the author of A Caution to Gentlemen who use Sheridan’s Dictionary seems to be quite certain that a diphthongal pronunciation surfaces. He comments (Anon. 1790: 21–22) that there is “a corruption which pervades every page of Mr. Sheridan’s Dictionary”, the monophthongal pronunciation of the ai graph: “People of rank and education always pronounce – AI – as a diphthong […]”. Batchelor’s (1809: 44) system of diphthongal representation through the addition of “y consonant” to simple vowel graphs also suggests too that diphthongal status of vowels in maid/take items has been of some temporal long-standing, missed through the lack of careful observation of earlier commentators: “(iy) as in tree; (ey) as in hey; (uy) as in buy; (oy) as in boy and (ay) as in ay”. Yet his contemporary Smart (1810: 94) sees the vowel in fate as monophthongal, while twenty six years later he seems to have changed his mind, proposing a diphthongal quality for the vowel space in items like gate, gait, and pay. Under his Principles of Pronunciation he asserts: The English alphabetic accented a, in the mouth of a well-educated Londoner, is not exactly the sound which a French mouth utters either in fe´e, or in feˆte, being not so narrow as the former, nor so broad as the latter. Moreover, it is not quite simple, but finishes more slenderly than it begins, tapering, so to speak, towards the sound of e1 […] This tapering off into e1 cannot be heard in the unaccented alphabetic a1 owing to its shorter quantity [i.e. in items like aerial and retail: CJ] (Smart 1836: iv).
But perhaps the more common observation on mid vowel diphthongization is, unsurprisingly, when it appears in pre-[r] environments, well illustrated by Cooley’s (1861: xvi) observation that a diphthongal status of the vowel in words like bare, fare, and pain is the norm: “its sound is so lengthened by the guttural vibration of the r, and the ‘radical’ and ‘vanishing’ movements of the voice become so apparent, that though organically the same as the a in fate, the effect on the ear is more or less diphthongal”. Ellis (1869: 294) claims that “At the present moment the (ee, oo) of the South of England are actually changing into (ei, ou)”. Again, in his On Paleotype, under his italicized ee, he comments: “English ale [eel] (eeil), long of [e], French fe´e (fee) […] Usually replaced by (eei) or (eei) in England”, while In the North of England, in France, and Germany, no difficulty is felt in prolonging the pure sounds of (ee) and (oo), but in the South of England persons have in general such a habit of raising the tongue slightly after the sound of (ee), and both raising the tongue and partly closing the lips after the sound of (oo), that these sounds are converted into the diphthongs (ee’j, oo’w), or (eei, oou) where the (ee, oo) parts are long and strongly
833
834
VI Late Modern English marked, and the (i, u) terminals are very brief and lightly touched but still perceptible, so that a complete diphthong results, which however is disowned by many orthoepists and is not intended by the speaker (Ellis 1869: 234).
Ellis admits to difficulty in the precise nature of the diphthongal transition: “I seldom or never say (eei) or (ei) ending with a perfect (i) […] at the most, I seem to reach (ee+e1), shewing a glide, and that in the process of ‘vanishing’ the force of the voice decreases so much that it is very difficult to say what sound is produced” (Ellis 1869: 234). His reluctance to accept the widespread use of the diphthongal forms owes something to the set of speakers who appear most to use them (Ellis 1869: 294): “At the present moment the (ee, oo) of the South of England are actually changing into (ei, ou), and these sounds have been developed by the less educated, and therefore more advanced speakers, the more educated and therefore less advanced having only reached (eei, oou) although many of them are not conscious of saying anything but (ee, oo)”. Sweet (1877: 70) has no doubt that an [eɪ] diphthong is involved, and even sees a “near” MAY/MY merger: “Thus the first element of ‘long a’ in English, as in take, is generally (e) or (e), but in Broad Cockney pronunciation it is (æ), and the resulting diphthong is not only heard as belonging to the (ai) type, but actually passes over into it, the first element becoming the mid- mixed (eh), as in the ordinary pronunciation of ‘eye’ ”. Although there is a small amount of evidence from commentators in the late 18th century that mid back vowels undergo a similar process of diphthongization to something like [oʊ], the development is almost entirely context sensitive – occurring in pre-sonorant contexts as a kind of “breaking” (Horn and Lehnert 1954, Vol. 1: 329– 330). It is really only with Smart that we see what might be a context-free development of this kind. Describing his “long o”, he claims that it is “apt to contract toward the end, finishing almost as oo in too”. That this diphthongal usage is, once more, generally viewed as non-prestigious seems to be supported by his observation on the ou, ow digraphs (Smart 1836: vii): “It is true that the same letters are sometimes sounded o¯ or o1; as in soul, blow and follow; but in this case the proper pronunciation will be indicated by omitting the w, or else marking it as silent”. Ellis explains that: As regards my own pronunciation, I feel that in know, sow etc., regularly, and in no, so, etc., often, I make this labial change, indicated by (oo’w).Wherein does this sound consist? In really raising the back of the tongue to the (u) position, and producing (oou) or (o´ou)? Or in merely further closing or “rounding” the mouth to the (u) degree, thus (o´o-ou)? […] There is no intentional diphthong, but a diphthong results so markedly, especially when the sound is forcibly uttered, that I have often been puzzled and could not tell whether know, sow serere; no, so; or now, sow, sus, were intended; I heard (no´u, so´u). But these are exaggerations, and I believe by no means common among educated speakers. Whether they will prevail or not in a hundred years, those persons who then hunt out these pages will be best able to determine (Ellis 1874: 1152).
Ellis shows a distaste for mid vowel diphthongal outputs (“The sound (bo´ut) is not only strange to me, but disagreeable to my ear and troublesome to my tongue”) and he is unprepared to accept Bell’s widespread use of diphthongized mid vowel forms: “Mr. M. Bell’s consistent use of (e´i, o´u) as the only received pronunciation thoroughly disagrees with my own observations, but if orthoepists of repute inculcate such sounds, for which a tendency already exists, their future prevalence is tolerably secured” (Ellis 1874: 1152).
52 Late Modern English: Phonology Eighteenth century comment manifests considerable controversy (or, perhaps, observational confusion) over the precise phonetic nature of “long u”, the descendant of Middle English [eu] and [ɛu], especially as to whether it is diphthongal (with yod onset) or monophthongal. Greenwood (1711), for instance, normally a slavish copier of Gildon and Brightland, seems to suggest unequivocally that long (u) is diphthongal. His initial description quite clearly echoes that of Gildon and Brightland (1711: 242): “The long Vowel U is pronounc’d like the French U, with a small or slender Sound; as lute, mute, muse, cure, &c” and he recommends marking it off diacritically from short and other versions of (u) by “a Point or Accent plac’d at the Top of U”. However, Greenwood (1711: 246) adds the important observation that this small or slender sound is one “as it were made up of I and W”. Indeed, he goes on to claim that “Eu, ew, eau, are Sounded by clear e and w; or rather u long. As in Neuter, few, Beauty, &c. But some pronounce them more sharp, as if they were to be written Niewter, fiew, bieuty, or niwter, fiw, biwty &c. especially in the Words new, few, snew. But the first way of pronouncing them is the better”. By mid-century, Flint sees the “long u” as diphthongal, using a transcription in iou¨, described as “iou Fr bref ”, exemplified by europe/Europe/you¨rop; beauty/beaute´/biou¨ti; dew/rose´e/diou¨ etc., according [ju] pronuncations to items like plume, rheum, lewd but not in blue, true, accrue, fruit, bruit, recruit, bruise. “In addition he refers to the wavering usage in blew, brew, clew, creew, drew, grew, slew, screw, threw, brewer, which were obviously pronounced with and without [ j], the latter variant being the more common one” (Ko¨keritz 1944: 109). By the 19th century, diphthongal values for “long u” are becoming ever more recognized. Batchelor almost always treats “long u” as a diphthong composed of an initial y consonant, followed by his long oo. Savage (1833: xxvii) is uncompromisingly hostile to such diphthongal outputs: “The following belong to the most wretched absurdities of the stage, and even there to the most low and uneducated. The dyuke made a nyew tyune upon the tyulip covered with dyew, and brought it to the tyutor on tyuesday”.
3 Consonantal developments 3.1
CH-ING:
[ti]/[tj]/[ʧ ] and [di]/[dj]/[ʤ] alternations
The changes under discussion here typically involve voiced and voiceless alveolar obstruents in syllable-initial position, preceding a vowel or semivowel segment with a high F2 (palatal) configuration – [i]/[ j]. In such palatal environments the change itself seems to have a diachronic sequence something like [ti] → [tj] → [ʧ ] → [ ʃ ], alongside a corresponding [dj] → [ʤ] → [ʒ]. For instance, Gildon and Brightland (1711: 54–55): “the English ( j) Consonant, or soft (g), or (dg), are compounded of (d) and (y), as is plain from Jar, joy, gentle, lodging, which sound Dyar, dyoy, dyentle, lodying, &c”. For them such a change seems to be a negative development and to constitute: “A certain Abuse to give the Sound of (s) to (c), before an (e) or (i), and of pronouncing (g) before the same Vowels, otherwise than before the others, of having soften’d the (s) between two Vowels, and of giving the (t) the Sound of (s) before (i), followed by another Vowel, as Gratia, Action, Diction, &c”. They provide extensive lists of items where [t] has been fully fricativized to [ ʃ ] in ian environments, with the sound of (shal ) given for (ti) in Credential, Essential, Nuptial, Impartial; (shan) for cian in Grecian, Logician, Magician; (shate) for ciate in Gratiate, expatiate, negotiate (“except
835
836
VI Late Modern English emaciate, Associate, Nauseate”); (shent) for cient in Ancient, Proficient, for tient in Patient, Impatient, for scient in Omniscient, and many others. Watts (1721: 130) too sees this change as non-prestigious usage: “I have therefore chosen out chiefly those words which are written different from their common and frequent Pronunciation in the City of London, especially among the Vulgar”. Among such items he selects: ainchunt ‘ancient’, conshunce ‘conscience’, conshenshus ‘conscientious’. Nares (1784: 129) describes the [ti]→[ʧ ] innovation as a “pronunciation which has been creeping in upon us very perceptibly for some years past”, but even only a few years later, Coote (1788) is sceptical concerning the propriety of such developments: “T when it precedes ion […] has the sound of s (unless the t follow s or x); as creation, nation, repletion, transition, notion, ablution. The sound of the ti in these words is corrupted by custom and rapidity of utterance into sh […]”. Almost all 19th century observers have some (usually extended) comment to make on this phenomenon. From Savage’s materials we get again what is a much more complex picture, one which shows much evidence of lexical diffusion. Lack of palatalization to some form of [ʧ ]\[ʤ] he regards as a “low” vulgarism – recording what he regards as stigmatized forms in: fixters ‘fixtures’; feeturs ‘features’; juncture ‘juncture’; venter ‘venture’, the orthoepy form of all of which Savage indicates with a tch graph. Certainly his evidence suggests that distribution of the palatalizing innovation is lexically motivated as much as anything else. For Ellis (1869: 203) this palatalization is no more than a “tendency”: “T,D have now a tendency, ignored by most orthoepists, under particular circumstances to pass into (tsh, dzh); thus nature, verdure are, perhaps most frequently, pronounced (neetshɹ, vɹdzh […] It is a fashion in modern English to resist, or to believe that we resist, this tendency in the especial case of -ture and -dure, but we have given in to it completely in -tion”.
3.2 Glide insertion Insertion of [ j] and [w] word initially after obstruents is only very occasionally recorded by observers pre-1750. Brightland and Gildon (1711: 55) recognize the phenomenon, attributing it to some kind of assimilatory process: “when a Palatal Vowel follows; for can, get, begin, &c. sound as if they were written cyan, gyet, begyn, &c. […] Pot, Boy, boil, &c which are sounded as if spelt thus, Pwot, Bwoy, bwoil, &c. but this is not always done, nor by all Men”. In the second half of the 18th century, however, the phenomenon is fairly widely recorded. No less than Walker (1791: 13) highlights the fact that “When a is preceded by the gutterals, the hard g or c, it is, in polite pronunciation, softened by the intervention of a sound like e, so that card, cart, guard, regard, are pronounced like ke-ard, ke-art, ghe-ard, re-ghe-ard. This sound of the a is taken notice of in Steele’s Grammar […] which proves it is not the offspring of the present day”. Again Walker (1791: 21) records how sky, kind, guide, guise, disguise, guile, beguile, mankind “are pronounced as if written skey, ke-ind, gue-ise, disgue-ise, gue-ile, begue-ile, manke-ind”. But for Nares (1784) “ky-ind for kind is a monster of pronunciation, heard only on our stage”. This phenomenon continues to be well represented in the discussions of 19th century observers. Batchelor (1809: 59–60), appealing to ease of articulation criteria, notes how “(y) is often subjoined to the guttural consonants (c) (g), when a palatine vowel follows; for, can, get, begin, &c. sound as if they were written cyan, gyet, begyn, &c. for the
52 Late Modern English: Phonology tongue can scarce pass from these guttural consonants to form the palatine vowels, but it must pronounce (y)”. He is careful to constrain the [j] insertion to such contexts, noting how “it is not so before the other vowels; as, in call, gall, go, gun, goose, come, &c.”. Ellis (1869: 206, 600) suggests that [j] insertion “is now antiquated in English” but that while “the custom is now dying out”, although “antiquated” it is “still heard”. Yet he records his own pronunciation of the item girl as (gjəəl), something like [ɡjɜɜl] (Ellis 1874: 1156).
3.3 H dropping and adding and [hw-]/[w-] alternations There is almost total silence on this phenomenon by most observers between 1700 and 1750, and what little comment there is comes almost entirely free of sociolinguistic constraint. Gildon and Brightland’s (1711: 41) observations on the matter are typically brief, commenting on [h] that: “’Tis indeed sometimes near silent, as in honour, hour, &c. but so are many other Consonants in particular positions”, and we can only conclude in this area is that syllable and word initial [h]-loss was not a salient characteristic of the phonology of English of this period. However, later in the century, [h]-loss is recorded by most commentators, typically Nares (1784: 26), who notes how the sound “is irregular only in being sometimes without effect; as in these initials, heir, honest, hospital, herb, hour, humour, hostler. In herbage I think it is usually pronounced, though suppressed in herb: nor is it dropped in horal, horary, &c. though it is in hour, the origin of which is the same”. Walker (1791: 46) too notes that “At the beginning of words, it is always sounded, except in heir, heiress, honest, honestly, honour, honourable, herb, herbage, hospital, hostler, hour, humble, humour, humourous, humoursome”. It is important to notice how, for Walker at least, this is not an unconstrained phenomenon, but one restricted to a specific lexical set to such an extent that “there are so very few words in the language where the initial h is sunk, we may select these from the rest”. The Scot Sylvester Douglas describes [h] effacement (and addition) as a “most capricious defect” in some English individuals, one which is apparently randomly spread throughout the lexicon as well as inter-regionally, some people pronouncing the “h in as complete a manner as other people, in words where it should be mute or is not written”, while “I know a hair-dresser who has this singularity of pronounciation, and who often lays it down as a maxim to his customers, that nothing is so destructive to the air, as exposing it too much to the hair” (Douglas 1779: 128). Douglas presents a picture of wholesale lexical diffusion with insufficient data to point to real conditioning factors or lexical trends in either the Pure or the Scotch dialect, a characteristic of several other observers, notably Kenrick: The H indeed is sometimes totally mute […] and this is likewise the case with this letter both in French and in English; it being pronounced in both languages in the same manner when audible, and both frequently, and to all appearance arbitrarily, mute: so that the words, in which it is audible, and in which silent, can be known in both languages only by practice (Kenrick 1784: 45).
Still, compared with the situation in the following century (Mugglestone 1995: 107–150, 282–288), any adverse sociophonetic consequences of [h]-dropping and adding are, if anything, relatively unstressed by late 18th century observers. Even Walker, who sees the phenomenon as one of the faults of Londoners (although he does see it as more reprehensible than others, notably [w] for [hw] word initially) it is a “vice”, a “worse habit”
837
838
VI Late Modern English he claims to be particularly common among children, but not one which merits the fullest level of condemnation he can summon for other “vulgarisms”. Vocabulary (Walker 1797) too shows only muted criticism of unetymological [h] usage, under HOSPITAL: “Mr. Sheridan and Mr. Walker sound the h in hospitable, hospitably, and hospitality; and suppress it in this word, which is the best usage; though we certainly often hear the h pronounced in hospital also, but improperly”. For Ellis (1869: 221) “At the present day great strictness in pronouncing h is demanded as a test of education and position in society”, noting that thirty years earlier Smart had allowed [h]-loss to a small lexical set which included heir, honest, honour, hostler, hour, humble, and honour, although for Ellis the last two are in his time generally [h]-full, with any attempt at [h] suppression interpreted as “social suicide”. Yet Ellis (1869: 223) is somewhat reluctantly forced to concede that some kind of ongoing change from below is actively in progress, one perhaps most closely associated with the newly aspirant lower middle classes of his day: “it must be owned that very large masses of the people, even of those tolerably educated and dressed in silk and broad cloth, agree with the French, Italians, Spaniards, and Greeks, in not pronouncing the letter H”. He stresses how “The Scotch never omit or insert it”; “No Scotsmen omit the aspirate”, while “The Germans are equally strict” (Ellis 1869: 222, 598). Late Modern commentators also observe a word-initial [w]/[hw] alternation, one still evident in the modern language; compare Standard RP [wɑɪt] “white” with colloquial Scots [hwʌɪʔ]. There is a similar regional contrast in the late 18th century record as well, although there is much to suggest from English observers that the alternation was also to be found south of the border with, possibly, evidence that [hw] initial types were being lost there. Like many observers, William Smith (1795: li–lii) claims that “WH, going before any vowel, except o, forms a double consonant, including the powers of h and w; Ex. What, whale, why”. Walker stands out by viewing any kind of [hw-] to [w-] change with concern: This letter [h: CJ] is often sunk after w, particularly in the capital, where we do not find the least distinction of sound between while and wile, whet and wet, where and were. Trifling as this difference may appear at first sight, it tends greatly to weaken and impoverish the pronunciation, as well as sometimes to confound words of a very different meaning […] we ought to breathe forcibly before we pronounce the w, as if the words [what, while: CJ] were written hoo-at, hoo-ille, &c. and then we shall avoid that feeble, cockney pronunciation, which is so disagreeable to a correct ear (Walker 1791: 46).
Such a phenomenon he lists as the Third Fault of Londoners: “The aspirate h is often sunk, particularly in the capital, where we do not find the least distinction of sound between while and wile, whet and wet, where and were, &c” (Walker 1791: xiii), a fault which for him is only superceded in awfulness by the loss of syllable initial [h] in words like heir, herb, honest, etc. Nineteenth century observers too record [hw-]/[w-] alternation, Batchelor (1809: 105) seeing it as a provincial deviation: “In such words as begin with wh, as when, the h ought to be softly pronounced before the w; as (hwen), (hwot); &c; but the h is, in such cases, constantly and entirely omitted among the peasantry in Bedfordshire”. Smart (1810: ix) sees the aspiration as always present: “w is aspirated in wheat, whig &c. which are pronounced hweat, hwig, &c” and again (1836: 180): “both letters should be heard” in
52 Late Modern English: Phonology whale (hoo-ale); what (hoo-at), when (hoo-en). The Vulgarities of Speech Corrected (Anon. 1826: 258) is quite explicit in listing Wat ‘what’, weet ‘wheat’, wen ‘when’, wist ‘whist’, wite ‘white’, and many others as “vulgar sounds”, the “correct” versions being hwot, hwen, hwist, and so on. While Ellis (1869: 188) sees (wh) as “uncertain in the south”, he suggests some ambivalence in the social significance of the alternation: “Although in London and the south of England (wh) is seldom pronounced, so that (wAt) is the usual sound for both Wat and what, yet to write wot for what is thought to indicate a bad vulgar pronunciation” (Ellis 1869: 573), while (whoo, whuu) for who, “is heard from elderly provincials” (Ellis 1869: 580).
3.4 Syllable final [r] loss and insertion Dobson’s (1968: 992–993) view that “there is no evidence at all of the [Standard English] vocalization and loss of [r] in stressed syllables in any of [the] 15th to 18th century sources which are alleged to show it” seems to be borne out from a close inspection of the early 18th century materials (Hill 1940). Greenwood (1711) and Mattaire (1712) have nothing at all to say on the phenomenon of [r]-loss, while the near-alike lists of Owen (1732) and Watts (1721) only show the stereotypical Harsh/Hash and Marsh/Mash alternants, although the former’s list (Owen 1732: 105) also records sasnet for ‘sarsenet’. It is only Mather Flint who seems to record the loss of post-vocalic [r] as an observable, active phonological process, suggesting that the sound is often pronounced “more softly” (adoucir) or “less distinctly” (foiblement) than its French equivalent. In items like hard, harm, Flint observes how “A suivi de r est un peu long sans eˆtre ouvert, & l’r est prononce moins rudemont qu’en franc¸ois” (Ko¨keritz 1944: 11). Not only that, but Flint seems also to suggest that the loss of [r] postvocalically is, in fact, some kind of [r] vocalization, which brings with it a lengthening of the preceding vowel segment: “l’r devant une consonne est fort adouci, presque muet & rend un peu longue la voyelle qui le pre´cede, barb, guard, arm, tarn” (Ko¨keritz 1944: 41). An observation like this might point rather to a change from a trill [r] to an approximant in [ɹ], as well as to a full effacement of the segment. Flint’s list of items showing what appear to be [r] “weakening” or effacement is quite extensive, occurring mainly in those contexts where the [r] is the first element of a consonantal cluster such as [rd], [rt]; [rθ], [rs], [rʧ ], [rʤ], [rl], [rm], and [rn], clusters traditionally associated with vowel lengthening in the history of English phonology. Some examples he cites (where he shows [r] loss/weakening by italicization) include: hard, regard, retard, award, reward, wizard, bastard, vineyard, third, bird, yard, lord. Watts (1721: 126–131) sees [r] loss as “the dialect or corrupt Speech that obtains in several Counties of England” especially those which are frequent in London “among the Vulgar”. His list includes examples such as fust ‘first’, nus ‘nurse’, and pus ‘purse’. Intrusive [r] phenomena are, not unexpectedly, equally rare, with only Bailey (1726: 58) and Jones (1701: 91) showing handkercher ‘handkerchief ”, and Tuite (1726: 37) claiming that “curfew […] is pronounc’d curfer”, although in both instances something other than [r]-insertion may be involved. For Ellis (1869: 196) what appears to be “vocalisation” is a major characteristic of his contemporary phonology (Ellis 1874: 1153): “(park, kart) with a genuine short a, and trilled r sound to me thoroughly un-English, and (park, kart) are either foreignisms or Northumbrianisms”. He equates the process directly with the vocalization of syllable-final [l]. For Ellis, syllable final [r], especially when preceding a consonant “is a
839
840
VI Late Modern English vocal murmur, differing very slightly from (ə)”, a sound he represents throughout by (ɹ), claiming (Ellis 1869: 196) that “in the mouths of by far the greater number of speakers in the South of England the absorption of the (ɹ) is as complete as the absorption of the (l) in talk, walk, psalm, where its has also left its mark on the preceding vowel”. The (ɹ) is an “indistinct murmur, differing from (‘l) by not having any contact between the tongue and the palate, but similar to it, in absorbing a variety of other vowels” (Ellis 1869: 197). Smart (1810: 107–108) records several instances of [r]-insertion: “There is a cockney pronunciation of the following, and other, words in which the sound a4 occurs, and which consists in pronouncing r after the sound, though this letter is not present. Such a blemish must be carefully avoided”. He claims that items like jaw, paw, law, thaw, claw, law, gnaw, withdraw, bawl, and straw, can appear with such an unetymological postvocalic [r]. Batchelor (1810: 151) too notes this intrusion occurring wordfinally after a vowel in items like fellow, window, willow, noting that: “The unprotracted, but open o1, in these words, is often corrupted by the vulgar into er. This blemish must be carefully avoided”. Similar observations, often accompanied by long lists of examples, are given considerable prominence in works like The Vulgarities of Speech Corrected (1826), Jackson’s (1830) Popular Errors, Mistakes of Daily Occurrence (Anon. 1885), and Errors in Speech and Writing Corrected (Anon. 1817: 7) where, under the Idea entry, we find the observation: “IDEAR, for Idea. Many people are guilty of this error, when the following word begins with a vowel: as, I have not the least idear of it. For the same reason, they sound an r at the end of all Christian names ending in a, as, Is Mariar out? Is Lousiar at home? Great pains should be taken to avoid this error”. The phenomenon takes on almost mythic status in the much published Poor Letter H and Poor Letter R pamphlets, where [r]-insertion is likened to “the murdering of the Queen’s English” and should be protested against “even were it only on the grounds of philanthropy”.
4 References Anon. 1790. A Caution to Gentlemen who use Sheridan’s Dictionary. 3rd edn. London: G. Bourne and R. & T. Turner. Anon. 1817. Errors in Speech and Writing Corrected. Anon. 1740. The Irish Spelling Book. (English Linguistics 1500–1800, ed. by R. C. Alston, 199.) Menston: The Scolar Press, 1969. Anon. 1885. Mistakes of Daily Occurrence of Speaking, Writing, and Pronunciation Corrected. London: n.p. Anon. 1797. Vocabulary of Such Words in the English Language as are of Dubious or Unsettled Pronunciation. London: Printed for F. and C. Rivington, etc. Anon. 1826. The Vulgarities of Speech Corrected. London: Bullcock. Bailey, Nathan. 1726. An Introduction to the English Tongue. London: Cox. Batchelor, Thomas. 1809. See Zetterson (ed.), 1974. Beal, Joan C. 1999. English Pronunciation in the Eighteenth Century. Thomas Spence’s “Grand Repository of the English Language”. Oxford: Clarendon. Beal, Joan C. 2000. HappY-tensing: A recent innovation? In: Ricardo Bermu´dez-Otero, David Denison, Richard M. Hogg, and C. M. McCully (eds.), Generative Theory and Corpus Studies: A Dialogue from 10 ICEHL, 483–497. Berlin/New York: Mouton de Gruyter Beal, Joan C. 2004. English in Modern Times. London: Arnold. Buchanan, John. 1770. A Plan for an English Grammar-School Education. London: n.p.
52 Late Modern English: Phonology Chambers, J. K. and Peter Trudgill. 1980. Dialectology. Cambridge: Cambridge University Press. Cooley, Arnold James. 1861. A Dictionary of the English Language. London: Chambers. Coote, Charles. 1788. Elements of the Grammar of the English Language. London: Dilly. Dobson, E. J. 1968. English Pronunciation 1500–1700. 2 vols. London: Oxford University Press. Douglas, James. 1740. See Holmberg, 1956. Douglas, Sylvester. 1779. See Jones (ed.), 1991. Ekwall, Eilbert. 1975. A History of Modern English Sounds and Inflections. Oxford: Blackwell. Ekwall, Eilbert (ed.). 1911. The Writing Scolars Companion. Halle: Niemeyer. Ellis, Alexander John. 1869, 1870, 1874, 1889. On Early English Pronunciation. Vols. I-V. London: Asher and Tru¨bner. Entick, John. 1795. The New Spelling Dictionary. London: Dilly. Fogg, Peter Walkden. 1792, 1796. Elementa Anglicana, or, The Principles of English Grammar Displayed and Exemplified, in a Method Entirely New. 2 vols. Stockport: Clarke. Gildon, Charles and John Brightland. 1711. A Grammar of the English Tongue. (English Linguistics 1500–1800, ed. by R. C. Alston, 25.) Menston: The Scolar Press, 1967. Greenwood, James. 1711. An Essay Towards a Practical English Grammar. (English Linguistics 1500–1800, ed. by R. C. Alston, 18.) Menston: The Scolar Press, 1967. Hammond, Samuel. 1744. The Young English Scolar’s Guide. London: Whitridge. Harland, S. 1719. The English Spelling Book. London: Taylor. Hill, Archibald A. 1940. Early loss of [r] before dentals. Publications of the Modern Language Association 55: 308–359. Holmberg, Bo¨rje. 1956. James Douglas on English Pronunciation c. 1740. Lund: Gleerup. Horn, Wilhelm and Martin Lehnert. 1954. Laut und Leben: Englische Laugeschichte der neueren Zeit, 1400–1950. 2 vols. Berlin: Deutscher Verlag. Jackson, George. 1830. Popular Errors in English Grammar, particularly in Pronunciation. London: Effingham Wilson. Johnston, William. 1764. A Pronouncing and Spelling Dictionary. (English Linguistics 1500–1800, ed. by R. C. Alston, 95.) Menston: Scolar Press, 1968. Jones, Charles. 1995. A Language Suppressed. Edinburgh: John Donald. Jones, Charles. 2003. English Pronunciation in the Eighteenth and Nineteenth Centuries. Basingstoke: Palgrave Jones, Charles (ed.). 1991. A Treatise on the Provincial Dialect of Scotland. Edinburgh: Edinburgh University Press. Jones, John. 1701. Practical Phonology. (English Linguistics 1500–1800, ed. by R. C. Alston, 167.) Menston: The Scolar Press, 1969. Kenrick, William. 1784. A Rhetorical Grammar of the English Language. (English Linguistics 1500–1800, ed. by R. C. Alston, 332.) Menston: The Scolar Press, 1972. Ko¨keritz, Helge. 1934–35. English pronunciation as described in shorthand systems of the seventeenth and eighteenth centuries. Studia Neophilologica 7: 73–146. Ko¨keritz, Helge. 1944. Mather Flint on Early Eighteenth Century Pronunciation. Skriflev Kungl. Humanistika Vetenskapssamfundet i Uppsala. Uppsala and Leipzig: Almqvist & Wiksells. Labov, William. 1994. Principles of Linguistic Change. Internal Factors. Cambridge, MA: Blackwell. Mattaire, Michael. 1712. The English Grammar. (English Linguistics 1500–1800, ed. by R. C. Alston, 6.) Menston: The Scolar Press, 1967. McKnight, George H. 1928. Modern English in the Making. New York: Dover. MacMahon, Michael K. C. 1998. Phonology. In: Suzanne Romaine (ed.), The Cambridge History of the English Language. Vol. IV. 1776–1997, 373–535. Cambridge: Cambridge University Press. Mitford, William. 1774. An Essay Upon the Harmony of Language. London: Scott. Mugglestone, Lynda. 1995. “Talking Proper”: The Rise of Accent as Social Symbol. Oxford: Clarendon. Nares, Robert. 1784. Elements of Orthoepy. (English Linguistics 1500–1800, ed. by R. C. Alston, 56.) Menston: The Scolar Press, 1968.
841
842
VI Late Modern English Owen, John. 1732. The Youth’s Instructor. (English Linguistics 1500–1800, ed. by R. C. Alston, 14.) Menston: The Scolar Press, 1967. Savage, W. H. 1833. The Vulgarities and Improprieties of the English Language. London: n.p. Scott, John. 1774. The School-Boy’s Sure Guide; or, Spelling and Reading Made Agreeable and Easy. Edinburgh: Gavin Alston. Smart, Benjamin Humphrey. 1810. A Practical Grammar of English Pronunciation on Plain and Recognised Principles. London: Richardson. Smart, Benjamin Humphrey. 1836. Walker Remodelled: A New Critical Pronouncing Dictionary of the English Language. London: Cadell. Smith, William. 1795. An Attempt to Render the Pronunciation of the English Language More Easy to Foreigners. London: Gillet. Sweet, Henry. 1877. A Handbook of Phonetics. Oxford: Clarendon Press. Tiffin, William. 1751. See Ko¨keritz, 1934–35: 87–98. Trudgill, Peter. 1986. Dialects in Contact. Oxford: Oxford University Press. Tuite, Thomas. 1726. The Oxford Spelling-Book (Being a Complete Introduction to English Orthography). (English Linguistics 1500–1800, ed. by R. C. Alston, 41.) Menston: The Scolar Press, 1967. Walker, John. 1791. A Critical Pronouncing Dictionary. (English Linguistics 1500–1800, ed. by R. C. Alston, 117.) Menston: The Scolar Press, 1968. Watts, Isaac. 1721. The Art of Reading. (English Linguistics 1500–1800, ed. by R. C. Alston, 333.) Menston: The Scolar Press, 1972. Wells, J. C. 1982. Accents of English. 3 vols. London: Cambridge University Press. Wyld, Henry Cecil. 1937. A Short History of English. London: John Murray. Wyld, Henry Cecil. 1953. A History of Modern Colloquial English. Blackwell: Oxford. Zetterson, Arne (ed.). 1974. A Critical Facsimile Edition of Thomas Batchelor. Part 1. Lund: Gleerup.
Charles Jones, Edinburgh (UK)
53 Late Modern English: Morphology 1 2 3 4 5 6 7
Introduction The verb The noun The adverb The adjective Summary References
Abstract The present chapter presents an overview of the latest, mainly corpus-based research on morphological and morphosyntactic changes in the areas of the VP (participial verb forms, the third person singular -s inflection, reflexive structures, -ing complements, the subjunctive, preterit and past participle forms), the NP (NPs preceding verbal gerunds, number concord, plural formation), the ADVP (temporal and sentence adverbs) and the ADJP (comparative formation). Highlighting British-American contrasts wherever Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 842–869
53 Late Modern English: Morphology possible the observed developments are related to functional constraints governing morphological variation and change, such as the principle of rhythmic alternation, the horror aequi principle, the complexity principle and the distance principle. What at first sight looks like divergent developments of individual morphological structures often turns out to form part of a larger development towards a new equilibrium governed by functionally motivated requirements.
1 Introduction In 1998, one of the most comprehensive accounts on Late Modern English (LModE) (1700–1900) morphology was published in Denison’s (1998) Cambridge History of the English Language chapter on “syntax”, “there not being enough in the period […] for a separate chapter on morphology” (Denison 1998: 94). Since then two developments in particular have provided a range of new insights. Firstly, the LModE period is the time-span in the history of English which hosts the largest and most comprehensive body of historical data, much of it available in computer-readable form. Given the recent boom in quantitative methods of research, this abundance of data can be optimally exploited, so that observations can move beyond the mere statement of the existence or non-existence of morphological items towards finer-grained analyses of the functional motivations underlying their trajectories of change. Secondly, Late Modern English is the period in the history of English in which British-American contrasts emerge, which provide an additional testing ground for monitoring morphological change during that period. Thus, many of the morphological and morphosyntactic changes characterizing this period turn out to be systematic and highly revealing in terms of functional motivations. The selection of topics discussed here is necessarily eclectic and largely motivated by the aim to add new insights from recent studies that have not yet been included or given centre stage in previous overviews of LModE morphology (such as the highly instructive chapters by Denison 1998; Go¨rlach 1999, 2001; Tieken-Boon van Ostade 2009). They reveal that Late Modern English can be regarded as a period characterized by a functionally-motivated division of labor. When LModE morphology and morphosyntax are addressed, this is often done in reference to a “confused situation” (Poldauf 1948: 240), by portraying this period as undergoing regularization processes that are triggered by standardization or by relating the high degree of simplification of morphology achieved by that time (cf. Denison 1998: 92). The present overview sets out to show that what at first sight look like divergent developments of individual morphological structures can occasionally be shown to form part of a larger development towards a new equilibrium governed by functionally motivated requirements.
2 The verb 2.1 Participial verb forms Unlike Early Modern English (EModE) and Present-day English (PDE), the LModE period witnesses extreme degrees of variation in the area of participial verb forms,
843
844
VI Late Modern English e.g. lighted vs. lit, knitted vs. knit (cf. Schlu¨ter 2009a). Fusion of the original -ed(e) suffix with stem-final -t had led to the emergence of variant participles. The LModE period is crucial for the decline of lighted, for example, by paving the ground for the PDE situation, in which lighted is largely ousted by lit (see Figure 53.1). In contrast, knitted extends its domain during Late Modern English rendering it the majority form in Present-day English. These divergent developments are most pronounced during the LModE period. It is only after this period that the system moves towards establishing a new majority form and hence a new equilibrium. 100% 114 80%
1229 1185
60%
lighted lit
40% 559
20%
148
23 0% EModE
LModE
PDE
100% 66 80% 115
137 60%
knitted knit
40% 70
42 20% 0 0% EModE
LModE
PDE
Figure 53.1: Participial variants lighted vs. lit and knitted vs. knit from Early Modern to Presentday English (based on Schlu¨ter 2009a: 113, 119)
While the use of participial forms marks the 18th and 19th centuries as a time of immense variability with rapid morphological change, Late Modern English also displays a highly systematic division of labor between monosyllabic lit and disyllabic lighted: the choice of either morphological variant is largely governed by rhythmic requirements that aim to create an alternating rhythm (by preventing stress clashes and lapses). Schlu¨ter (2009a) contrasts three syntactic environments: 1. attributive uses (a lı´ghted ca´ndle) 2. complex attributive uses (an u´nlit sı´de cavern) 3. non-attributive uses (lı´t his pı´pe) (cf. Schlu¨ter 2009a: 113–114)
53 Late Modern English: Morphology
845
Lighted is particularly frequent in attributive position before an initially stressed noun, where -ed can serve as a rhythmic buffer avoiding a stress clash. In both British and American Late Modern English the decline of disyllabic lighted is more advanced in syntactic environments that do not run the risk of creating a stress clash. % lighted 100%
74
208
63
208 80%
attributive complex attributive non-attributive
60% 446
40%
91
41 20% 125
38
0% *1800–1829
*1830 –1869
*1870 –1899
Figure 53.2: Participial variants lighted vs. lit in Late Modern BrE (N = 1294) (based on Schlu¨ter 2009a: 115)
Figure 53.2 shows this change towards the lit-variant in British English (BrE) for three syntactic environments. A parallel though slightly delayed development has also been observed for American English (AmE) (cf. Schlu¨ter 2009a: 115). Historically, this development marks a change towards morphological irregularization: Innovative forms have been established faster in contexts where they promote rhythmic alternation, while they have been established more slowly in contexts where they lead to objectionable rhythmic constellations. Conversely, obsolescent forms have been given up more reluctantly in contexts where they help preserve an alternating rhythm, and have been given up more readily where they violate this universal principle (Schlu¨ter 2009a: 128).
In the context of discussions on the influence of standardization, prescriptivism, and codification on 18th century English, the morphological development described here is particularly revealing. After all, the opposing trajectories of change observed by Schlu¨ter (2009a) and shown in (1) for (1)
lighted > lit irregularization more advanced in BrE knit > knitted regularization more advanced in AmE
are unlikely to be the product of prescriptive influence. But while the participial forms of the two verbs appear to develop in opposite directions establishing lit and knitted as majority forms, the variability is in itself optimally exploited by creating a prosodically-motivated diversification of functions.
846
VI Late Modern English
2.2 Completion of the spread of the third person singular -s inflection Late Modern English sees the completion of a process which was largely but not entirely finalized in Early Modern English, i.e. the ousting of the southern 3P SG -th morpheme by northern -s. While this process affects most verb forms in the EModE period, high-frequency verbs such as hath, doth, and saith preserve the northern forms well into LModE, where they are finally completely supplanted by -s forms. Another factor operative in this transition from -th > -s is national variety, with the replacement process being more advanced in American English than in British English, except for verbs already ending in certain sibilants (e.g. riseth; cf. Kyto¨ 1993: 132).
2.3 Reflexive structures Research on Late Modern English reflexives reveals tremendous changes. While the reflexive has largely ousted personal pronouns in coindexical use by the EModE period, LModE witnesses a reversal of its expansion, shown in (2). (2)
I washed me > I washed myself > I washed Ø ME EModE LModE
The use of the reflexive decreases both in terms of type frequency (i.e. the number of verbs it can be used with) and token frequency (i.e. the number of occurrences). Rivaling structures oust the reflexive in certain functional domains, and this replacement process is particularly advanced in AmE (cf. Rohdenburg 2009a: 180). The competing strategies operative in the functional domain of the reflexive are given below in examples (3)–(6). Section 2.3.1 addresses each replacement process in turn. (3)
zero forms So I dressed Ø at leisure, and then went to Youwarkee, and waked her; (1752 Fielding, Amelia)
(4)
particle verbs But you’ll have to brace up. There’s going to be a marriage – ; (1893 Stevenson, The Beach of Falesa)
(5)
way-constructions yet I am not unknown to the African part of the Macrocosme, where my single Sword hath eaten its way through thousands, and hath afterwards drank it self into a Surfeit, with the blood of those Hell-dyed Infidels. (1668 Head and Kirkman, The English Rogue)
(6)
combinations of particle and reflexive “[…] Pray excuse my intrusion. Many thanks. Good day!” He bowed himself out; and Mrs. Sparsit, hiding in the window-curtain, saw him languishing down the street (1854 Dickens, Hard Times)
53 Late Modern English: Morphology
847
2.3.1 Replacement of the reflexive by zero forms The variation between the reflexive and zero forms can be illustrated as follows, (7): (7)
a. The Thames empties itself into the North Sea. b. The Thames empties Ø into the North Sea. (Rohdenburg 2009a: 167)
British-American contrasts in the use of the reflexive were first noted by Jespersen (1961 [1928]: 331). Empirical support has recently been provided by Rohdenburg (2009a). Reflexive 100%
2
8
1
4
80%
37
25 13
60% 82 40%
oversleep empty keep from
20%
6
0% *1460–1700
*17th/18th cent
*19th cent
Figure 53.3: Diachronic development of the use of the reflexive as opposed to the zero-variant in BrE (based on Rohdenburg 2009a: 168–170; years marked by an asterisk indicate birth dates of the authors rather than publication dates)
Figure 53.3 illustrates the decline of the reflexive to the benefit of the zero variant for three verbs in British prose fiction data. For the verb keep (oneself) from the zero-variant is already a fairly well-established option in the LModE period. Late Modern AmE is considerably more advanced in the replacement of reflexives by zero forms (see Figure 53.4). While in BrE empty oneself is still the rule for authors born in the 17th and 18th centuries, for AmE the reflexive is the exception. Additional studies on some 120 reflexive verbs indicate that at present “the erosion of the reflexive pronoun is continuing at a striking rate in AmE” (Rohdenburg 2009a: 172). As regards the pace of change, frequent verbs (such as undress oneself, hide oneself ) are affected more strongly and earlier than less entrenched verbs (such as brace oneself, ensconce oneself ) (cf. Rohdenburg 2009a: 173). What is more, if the verb is followed by a preposition, the omission of the reflexive is even more likely, as seen in (8) and (9): (8)
indulge
oneself
in something
(9)
indulge
Ø
in something
848
VI Late Modern English Reflexive 100% empty oversleep keep from
11 80%
60%
40%
5
22
20%
55
1 16 0% *17th/18th cent
*19th cent
Figure 53.4: Diachronic development of the use of the reflexive as opposed to the zero-variant in AmE (based on Rohdenburg 2009a: 168–170)
The decline of the reflexive, according to Rohdenburg (2009a: 174), can be related to the degree of argument complexity: if prepositional or infinitival complements are present, the reflexive is strongly dispreferred. In still other cases the reflexive tends to be inserted in accordance with the horror aequi Principle (cf. Rohdenburg 2003a: 236–242), which constrains the use of formally identical, adjacent grammatical elements. In this case the occurrence of two adjacent to-infinitives is avoided by retaining the reflexive. In general, the diachronic dimension of horror aequi implies that variants are retained considerably longer if they can serve to avoid the adjacency of two (near) identical grammatical elements, such as the infinitives in (10a). An alternative strategy in avoiding the use of two immediately adjacent to-infinitives is used in (10b), where the first to-infinitive is replaced by the -ing form. (10) a. They intended to organize themselves to defend their rights. b. They were organizing Ø to defend their rights. (based on Rohdenburg 2009a: 176) This apparently applies even though the to-infinitive in (10a) is not a complement of organize, but introduces a purpose clause. Horror aequi can thus extend to non-arguments as well as arguments.
2.3.2 Replacement of the reflexive by particles The decreasing use of the reflexive (11) is also fostered by an enhanced number of particles (12) or hybrid combinations of reflexive plus particle, as in (13). (11)
reflexive For some undefined reason she felt as if she could not. Yet she braced herself for her lonely walk with something of an effort. (1897 Coleridge, The King with Two Faces)
53 Late Modern English: Morphology
849
(12)
particle Prepare yourself. I speak positively. You have to brace up for one sharp twitch – ; the woman’s portion! as Natata says (1891 Meredith, One of Our Conquerors)
(13)
reflexive + particle but as it was inevitable, she braced herself up for his inspection by closing her lips so as to make her mouth quite unemotional (1887 Hardy, The Woodlanders)
Analyses of 27 verbs permitting reflexives, particles, or hybrids show that a change in which particle verbs start supplanting historically older reflexive forms is beginning in LModE (see Figure 53.5). 100% 229
562
80% 1061 1026
reflexives particles reflexives + particles
60%
40%
535
20% 22 0%
21 1460–1700
169
59 38 1700 –1800
146 1800 –1900
165 1960 –1993
27 verbs: boost, bow, brace, burn, coil, curl, dress, ease, empty, fit, freshen, gear, heave, hire, jerk, launch, lock, open, prop, psyche, raise, rouse, smarten, straighten, stretch, work, wrap
Figure 53.5: Diachronic development of the competition between reflexives versus particles (N = 3619) (based on Mondorf 2010)
This replacement process gains momentum in the second half of the LModE period, when the particle even becomes the majority form. The role of the hybrid form (reflexive plus particle) has remained constant over the last 500 years.
2.3.3 Replacement of the reflexive by way-constructions The reflexive is also increasingly being ousted by way-constructions (illustrated in example 5 above). Having been around as early as the OE period, this construction has continually extended its territory both in terms of frequency and in terms of its range of applications (cf. Israel 1996: 227 for a comprehensive account of the grammaticalization of way-constructions). Historical corpora reveal that the 18th century marks a turning point during which the way-construction becomes the majority form with a
850
VI Late Modern English selection of ten verbs investigated in Mondorf (2011). The use of self erodes over time, while way has been gaining ground (see Figure 53.6).
100% 358
80% 48 60%
284
71
40%
57
206
27 way self
20%
95
0% 1460 –1700
1700 –1800
1800 –1900
1960 –1993
10 verbs: cut, drink, eat, fight, grope, hit, wind, work, worm, wriggle
Figure 53.6: Diachronic development of the competition between way vs. self (N = 1146) (based on Mondorf 2011)
Apart from these quantitative differences, we also observe an emergent division of labor: the way-construction is particularly successful in superseding the reflexive in concrete uses (gauged in terms of the noun in the directional), e.g. (14) and (15): (14)
concrete use And Ulfketyl worked hard and well, till a string of barges wound its way through the fens, laden with beeves and bread, and ale-barrels (1848 Bronte¨, The Tenant of Wildfell Hall )
(15)
abstract use She knew not why, she could not tell how the girl had wound herself round her maternal heart. (1862 Thackeray, The Adventures of Philip)
Conversely, the reflexive can to some extent stand its ground in abstract contexts. The availability of both variants is thus accompanied by an emergent semanticallymotivated division of labor: concrete uses are a marked domain of the way-construction, while abstract uses preserve a higher share of the older reflexive. In the long run, however, the way-construction is gradually extending its domain across the board, so that the reflexive is clearly on the retreat (cf. Mondorf 2011: 405).
53 Late Modern English: Morphology
2.3.4 Avoidance of reflexive verbs in Late Modern American English An interesting change concerning reflexives in AmE is a tendency towards an overall avoidance of verbs requiring a reflexive (cf. Rohdenburg 2009a: 179). In BrE this type of avoidance is far less pronounced. This distribution is in line with a general trend concerning British-American contrasts, in which AmE – when given the option – tends to avoid variants that exert a higher processing load to the benefit of those that are more economical in terms of “processing” (cf. Mondorf 2009: 97–99; Rohdenburg 2009a: 177–181). The origins of the trend towards the zero-variant can be traced at least as far back as the Late Modern English period. It has affected American English much faster and more extensively than British English. The majority of developments taking part in this development seem to have evolved during the present-day English period. In the case of verbs whose reflexive pronoun cannot be replaced by zero (without dramatic semantic changes), AmE has at least for something like fifty years led BrE in the decline of reflexive uses. This fact may be attributed to the stronger tendency of AmE […] to avoid comparatively complex and formal structures (Rohdenburg 2009a: 180).
We can summarize that Late Modern English witnesses drastic changes in the system of reflexives which lay the foundation for subsequent British-American differences in this area of morphosyntax.
2.4 The spread of -ing complements A morphosyntactic development that forms part of the so-called “Great Complement Shift” (cf. Rohdenburg 2006b: 144; Vosberg 2009: 213) is well observable in Late Modern English. It concerns the reorganization of the system of sentential complementation. While in the ME period that-clauses were increasingly replaced by to-infinitives, the infinitive in turn gave way to -ing complements in the LModE period (cf. Fischer 1997: 267–268; Fanego 1996: 77; Kjellmer 1980: 89–92; Vosberg 2003: 2009; Mair 2006). In a highly simplified fashion, the development might be sketched as follows: that-clauses
> that-clauses > that-clauses to-infinitives to-infinitives -ing complements
In Present-day English the to-infinitive is more often used to express purpose and future orientation, while the -ing complement is more closely associated with retrospective meanings, a distribution that does not yet hold as categorically in Late Modern English as it does today (examples from Vosberg 2003: 305, 306): (16)
-ing complement I remember hearing of a dispute between two youthful clerks, (1861 Meredith, Evan Harrington)
(17)
infinitival complement “The cheapest policy of Insurance I remember to have heard of!” said Adrian. (1859 Meredith, The Ordeal of Richard Feverel)
851
852
VI Late Modern English In these examples, the infinitival complement is still admissible with retrospective meanings. What is more, the development results in a tendency towards a functional split in non-canonical syntactic structures, such as extraction contexts (as in 17, where the of-complement has been moved to sentence-initial position), and non-extracted clauses (illustrated in 16 cf. Vosberg 2003: 307–308; Warner 1982; Mair 1993: 6 and Fanego 1996: 79): “In the case of infinitival or gerundial complement options, the infinitive will tend to be favored in environments where a complement of the subordinate clause is extracted (by topicalization, relativization, comparativization, or interrogation etc.) from its original position and crosses clause boundaries”(Vosberg 2003: 308). Assuming that extraction contexts require a higher processing effort Vosberg (2003) finds that the establishment of the historically more recent -ing complement is delayed in these extraction contexts. (For an explanation in terms of the Complexity Principle, see Vosberg 2003). The replacement of -ing complements by the to-infinitive is thus more advanced in syntactically complex environments (see Figure 53.7). % -ing
British English
American English
100%
100%
80%
80%
60%
60%
40%
40%
20%
20%
0%
0% *1728–1799 *1800 –1869 *1870 –1894 no extraction
*1728 –1799 *1800 –1869 *1870 –1894 extraction
Figure 53.7: The distribution of perfect infinitives vs. simple -ing forms involving past reference complementing the verb remember in BrE and AmE historical corpora (p < 0.1% *** for each period) (based on Vosberg 2003: 311–312)
Another construction taking part in this replacement of to-infinitives by gerundial complements is to lay claim. By the end of the LModE period the to-infinitive emerges as sentential complement which is increasingly replaced by the -ing form after the LModE period, shown in (18). (18)
It has afforded the Author great amusement and satisfaction, […] to learn from country friends […] that more than one Yorkshire schoolmaster lays claim to being the original of Mr. Squeers. (1839 Dickens, The Life and Adventures of Nicholas Nickleby)
While in the LModE period BrE has spearheaded the change towards -ing complementation with AmE following suit, in Present-day English the American standard has even overtaken BrE in supplanting the to-infinitive with to lay claim. However, the
53 Late Modern English: Morphology replacement process can be delayed in certain contexts where processing requirements call for the older and more explicit infinitival variant (cf. Vosberg 2009: 219–223). The reverse trend, i.e. a change from gerundial complementation (19) to infinitival complementation, is, however, observed for can’t stand (cf. Vosberg 2009: 224): (19)
’Pon my honour, I can’t stand seeing a whole family going to destruction! (1818 Ferrier, Marriage; Vosberg 2009: 224)
While the -ing form was used in the vast majority of cases at the end of Late Modern English, the to-infinitive has become the majority form in the late 20th century, especially in AmE (cf. Vosberg 2009: 224; Fanego 1996: 75–76; Rohdenburg 2006b; Milroy and Milroy 1993: 130).
2.5 The subjunctive While the English subjunctive has occasionally been considered to have become virtually extinct in the EModE period, it has since seen a significant revival, spearheaded by AmE.
2.5.1 The mandative subjunctive The mandative subjunctive is one of the best-known and most controversially discussed morphosyntactic changes in the history of English. The observation that it is presently more wide spread in AmE than in BrE occasionally led to the assumption that its frequent use in the American variety represents a conservatism. By contrast, recent empirical studies concur that the subjunctive had virtually become extinct in both varieties; rather than witnessing its delayed demise in AmE, we are observing its revival in ¨ vergaard 1995; Hundt 2009; AmE and – though at a slower pace – also in BrE (cf. O ¨ Kjellmer 2009; Schluter 2009b; Crawford 2009). The trajectory of change takes the form of a successive decline from Old English to Early Modern English, ranging from a relatively wide distribution in Old English, with competition between indicatives and modal periphrases (e.g. scolde + infinitive), via a reduction of formal marking in Middle English (when the indicative preterit plural -on and subjunctive preterit plural and past participle of strong verbs -en were fused and final unstressed -e was lost), to rare instances in Early Modern English. It is only at the end of the LModE period that the subjunctive re-established itself and “nothing less than a revolution took place” (Kjellmer 2009: 247). The competition between irrealis marking by means of modal periphrasis and the subjunctive is illustrated in the following example, (20): (20)
Every student should be emulously watchful that he do not diminish the stock of professional credit by his idleness; (1809 More, Coelebs in Search of a Wife)
In answer to the question of what caused the resurrection of the subjunctive, it is often argued that the subjunctive offers the advantage of marking irrealis, as does modal periphrasis. Marking this distinction was almost lost in BrE but not in AmE, which strongly exploited the option of marking irrealis by modal periphrasis. With a decline
853
854
VI Late Modern English in the use of certain modal verbs in AmE, a functional gap emerged that might have been filled by the subjunctive: “AmE simply substituted one marked form for another – and thereby turned a conservative trait into an innovation” (Schlu¨ter 2009b: 291; cf. Kjellmer 2009). For Late Modern English the rivalry between the subjunctive and modal periphrases with the verb demand in combination with the be-passive is illustrated by (21) and (22). (21)
modal periphrasis Still Captain Kearney, although a brave and prudent officer – ; one who calculated chances, and who would not risk his men without he deemed that necessity imperiously demanded that such should be done – ; (1834 Marryat, Peter Simple)
(22)
subjunctive we require and demand that the said noblepersons, namely, Cedric of Rotherwood, Rowena of Hargottstandstede, Athelstane of Coningsburgh, with their servants, cnichts, and followers, also the horses and mules, Jew and Jewess aforesaid, together with all goods and chattels to them pertaining, be, within an hour after the delivery hereof, delivered to us, (1820 Scott, Ivanhoe)
The trajectories of change reveal that the AmE lead in re-establishing the subjunctive at the expense of modal periphrastic constructions is well under way in the LModE period, while we merely observe the incipient stages for BrE. This development is one of the more spectacular changes during the LModE period (see Figure 53.8). As regards the question of whether the AmE use of the subjunctive spilled over to BrE, “[t]his modest rise of the passive subjunctive […] allows us to speculate that contrary to received opinion the subjunctive revival in British English might after all be a partially homegrown development” (Rohdenburg 2009b: 322). Subjunctive 100% 20/23
AmE BrE
80%
28/42
60%
40%
20%
3/10 0/13
8/26
7/39
2/9
1/31
0% *1650 –1799
*1800 –1830
*1830 –1869
1870 –1900
Figure 53.8: The distribution of subjunctives vs. modal periphrases in combinations of the verb demand with be-passives (based on Rohdenburg 2009b: 321)
53 Late Modern English: Morphology
855
2.5.2 The subjunctive in adverbial clauses In adverbial clauses introduced by if, before, whether, ere, unless, however, till, whatever, (al)though, except, whatsoever, whomsoever, lest, until, howsoever, whosoever, whoever, as if and so that, the subjunctive is on the retreat during the EModE and LModE periods (cf. Auer 2006: 44; see Figure 53.9). This group of clauses does not yet illustrate the reversal encountered for the mandative subjunctive. 100% subjunctive indicative modal periphrasis
80%
60%
40%
20%
0% 1570–1710
18th century
19th century
Figure 53.9: The distribution of subjunctives, indicatives and modal periphrases in adverbial clauses (based on Auer 2006: 44)
With lest, however, the subjunctive has become almost obligatory in Present-day AmE and it is also increasing its territory in BrE (cf. Rohdenburg and Schlu¨ter 2009: 410). The subjunctive with lest was still used in Early Modern English, then became obsolete in Late Modern English, and then experienced a revival in Present-day English (cf. Auer 2007: 165). As regards the conditional subjunctive introduced by on (the) condition (that), the subjunctive also undergoes a resurrection which starts in the 19th century. It was almost extinct between the 16th and 20th centuries, amounting to a miserly 1–3% of all occurrences, but has risen again to roughly 9% of all uses in Present-day English. In contradistinction to mandative clauses as well as other conditional clauses, the subjunctive thus represents a true newcomer in clauses introduced by the complex conjunction on condition, where it quickly imposed itself in twentieth-century AmE. In BrE, the rise of the subjunctive is only in its infancy in the late twentieth century (Schlu¨ter 2009b: 291).
The revival of the subjunctive is illustrated in Figure 53.10 (years marked by an asterisk indicate birth dates of the authors rather than publication dates).
856
VI Late Modern English 100%
80%
subjunctive ambiguous indicative modal
60%
40%
20%
60 – = 2 200 65 3 N
19
87 0 N –*1 = 1 89 1 9
*1
83 0 N –*1 = 1 86 08 9
*1
80 0 N –*1 = 1 82 39 9
*1
72 8 N –*1 = 1 79 09 9
*1
66 0 N –*1 = 1 75 19 2
*1
*1
46 0 N –*1 = 6 67 1 0
0%
Figure 53.10: The verbal syntagm in subordinate clauses dependent on (up)on (the) condition (that) in BrE (based on Schlu¨ter 2009b: 288)
As regards AmE, the incipient increase in the use of conditional subjunctives is observable considerably earlier, i.e. in the outgoing LModE period (see Figure 53.11). 100%
80%
subjunctive ambiguous indicative modal
60%
40%
20%
60 N –20 = 2 03 3
19
70 – N *18 = 8 99
*1 8
83 0 N –*1 = 6 86 6 9
*1
80 0 N –*1 = 1 82 03 9
*1
*1
72 8 N –*1 = 7 79 3 9
0%
Figure 53.11: The verbal syntagm in subordinate clauses dependent on (up)on (the) condition (that) in AmE (based on Schlu¨ter 2009b: 289)
53 Late Modern English: Morphology
857
The subjunctive has even become the majority form with on (the) condition (that) in Present-day AmE: [S]ince the mid nineteenth century, BrE has exhibited a pronounced tendency to abandon the irrealis (mandative) marking of the verbal paradigm which continued to be indicated by the use of modal auxiliaries in AmE. This difference has been argued to be at the bottom of the susceptibility of AmE to the subjunctive revival, while the concrete trigger seems to have been the parallel change in mandative clauses. The first signs of the comeback have been traced to the late nineteenth century, and the change has spread so rapidly in AmE of the early twentieth century that the subjunctive has been the majority variant since about 1910 (Schlu¨ter 2009b: 304).
2.6 Preterit and past participle forms The system of irregular past tense forms of weak verbs (e.g. burned vs. burnt, smelled vs. smelt, etc.) has seen dramatic changes throughout the history of English. The historically earlier -ed(e) suffix was followed by phonologically assimilated -t forms. As regards British-American differences, Lass (1999: 175) observes that irregular verbs “now generally keep the old /-d/ forms in the US (smelled, burned), while in BrE and the Southern Hemisphere Extraterritorial Englishes they have the newer /-t/”. While this at first sight suggests a development in the form of a so-called “colonial lag” or “extraterritorial conservatism” of American English, analyses of Late Modern English reveal that the irregular /-t/ forms are already the majority in 18th AmE (cf. Hundt 2009). The minority status of /-t/ forms in the 19th century thus reflects a reversal of the trend from /-d/ > /-t/ in the direction of /-t/ > /-d/. This reversal leads to an increase in regularization, since verbs such as smell, burn, etc. increase their share of smelled, burned at the expense of smelt or burnt . This regularization process is less advanced in 18th century AmE than in the British homeland, but it proceeds at a quicker rate, resulting in today’s greater extent of regularized -ed forms in AmE (cf. Hundt 2009: 20–27).
3 The noun 3.1 The rivalry between possessive verbal gerunds
NPs
and objective
NPs
preceding
Late Modern English just as Present-day English permits variation between the possessive and the objective form of an NP preceding a verbal gerund, as illustrated in the PDE example: (23)
a. She liked his being childlike with her. possessive more formal b. I liked him being out of control, trying objective less formal to grasp on to anything or anyone. (Corpus of Contemporary American English, Davies 2008–)
Though both variants are considered grammatically correct in Present-day English, the possessive is more closely associated with formal contexts (cf. Quirk et al. 1985: 1194).
858
VI Late Modern English The large extent of variation in Late Modern English is illustrated in the following text passage, in which both variants are used by one and the same author in almost identical contexts: (24)
“I don’t mind your calling me a clog, if only we were fastened together. But I do mind you calling me a donkey,” he replied. (1866 Gaskell, Wives and Daughters)
While 18th and 20th century prescriptivists aim to ban the objective, the debate concerning the acceptability of the possessive culminates in the 20th century with Jespersen (1926: 148, 150) branding Fowler (1925: 44) an “instinctive grammatical moralizer” (cf. Peters 2004: 229; Lyne 2006: 39). In the late 18th century, pronominal NPs are almost exclusively used with the possessive variant, while full NPs are equally divided among both variants (see Figure 53.12). The decrease of the possessive preceding verbal gerunds is well under way during the LModE period with full NPs, while pronominal NPs undergo a time-lagged decline in the 20th century. As regards British-American contrasts, the older possessive variant is still more frequently used in AmE (cf. Hudson 2003: 581; Rohdenburg and Schlu¨ter 2009: 418; for analyses of the impact of style, text type, syntactic position, etc., see Heyvaert et al. 2005; Lyne 2007). 100% 492
798
pronominal NP full NP
80%
60%
74 104
40% 61
20%
4 0% 1751–1800
1851–1900
Figure 53.12: Diachronic development of possessive (based on Lyne 2007: n.p.)
1960 –1993 NPs
preceding verbal gerunds (N = 2322)
3.2 Concord with collective nouns An increase in regularization that can also be described in terms of option-cutting or an emergent division of labor can be observed for concord with collective nouns. While Liedtke (1910: 180–181) reports a rising use of plural verb agreement from Old English to Early Modern English, Dekeyser (1975: 180–181) observes an apparent reversal, i.e. declining uses of the plural variant for the 19th century. Levin (2006: 304) links the reduction of plural verb agreement since the EModE period to an emergent
53 Late Modern English: Morphology
859
specialization: plural marking tends to become more frequent if the verb is more distant from the subject, as e.g. in (25) and (26) (25)
The club, blinded by the gargantuan deeds and magnanimity of a player easily worth three times his Essex salary, were deaf to the needs of others (based on Levin 2006: 340)
(26)
Our club is better than Baywatch (based on Levin 2006: 340)
In line with the complexity principle (Rohdenburg 2003a: 217) this suggests that the more explicit variant, in this case the variant containing an additional plural morpheme, comes to be used in cognitively more complex environments. According to Marckwardt (1958: 77), the use of plural verbs with collective nouns (e.g. crew, government, family) is an instance of American conservatism. Conversely, Hundt (2009: 28) and Levin (2001: 36) are able to show that AmE is more advanced than Present-day English as regards the historical trend in concord with collective nouns. As Figure 53.13 reveals, singular concord with collective nouns reaches a trough in the 19th century and increases again up to the 20th century, with AmE spearheading this development. Singular Concord 100%
80%
60%
40% BrE AmE
20%
0% 1600–1699
1700 –1799
1800 –1899
1900 –1990
Figure 53.13: Singular concord with collective nouns in BrE and AmE (N = 302) (based on Hundt 2009: 29)
A different case of number marking concerns number marking on the noun itself. Denison (1998: 97, 99) observes that acquaintance can be used as a collective noun with singular and plural concord in Late Modern English, while in Present-day English the noun is marked for number. (27)
I had been here but a little Time before I had a great many Acquaintance; among the Number of them was Signior Jaques Honorius Cittolini (1725 Haywood, Love in Excess)
860
VI Late Modern English Additional analyses of prose corpora confirm that the marked form acquaintances is a relatively recent LModE development, which gains a moderate foothold in the 19th century, while in Present-day English it is already far in excess of 30% (see Figure 53.14). 100% 1640
3152
2515
80% 243 60% acquaintance acquaintances
40%
127 20% 124
523
32 1460 –1699
1700 –1799
1800 –1869
0% 1960 –1993
Figure 53.14: Diachronic development of singular and plural forms of acquaintance (N = 8347)
Other nouns, such as accommodation(s), can be found to reduce explicit plural marking, as in (28). (28)
You must permit me, said she to Mr. Glanville, to intreat your noble Friend will accompany us to the Castle, where he will meet with better Accommodations than at any Inn he can find; (1751 Lennox, The Female Quixote)
These developments go hand in hand with a regularization in the system of number concord.
3.3 Plural formation 3.3.1 Plural formation: hoofs vs. hooves and scarfs vs. scarves Rather than promoting a trend towards regularization, a tendency towards irregularization originates in or around the LModE period. The plural form hooves as opposed to hoofs is not yet attested in OED citations of the 18th century but increases in use throughout the 19th and 20th centuries. Similarly, scarves continually extends its territory until becoming the majority form in Present-day English (cf. Rohdenburg 2003b: 276). The voiced variants thus become more frequently used in Late Modern English (see Figure 54.15). (For an explanation on the distribution of the variant forms in terms of iconicity, see Rohdenburg 2003b.)
53 Late Modern English: Morphology
861
100%
80%
60%
40%
hoofs scarfs hooves scarves
20%
0% 1700–1800
1800 –1900
1900 –1979
Figure 53.15: Diachronic development of the plural allomorphs of hoof and scarf in quotations from the Oxford English Dictionary (N = 211) (based on Rohdenburg 2003b: 276)
3.3.2 Plural formation: uninflected plurals after numerals Measure nouns, such as pound(s), ton(s), fathom(s), quid(s), are used variably with or without the -s plural inflection when preceded by a cardinal number expression in Late Modern English. The noun fathom illustrates that with some nouns the change from the zero-variant (fathom) to the -s-variant (fathoms) roughly starts and reaches completion during the LModE period (see Figure 53.16). Other examples displaying variation between zero-marking and the -s-inflection in varieties of English are foot, pound, shilling, mile, acre, year, month, week, hour, inch, gallon, load, ton, pair, penny, quid, etc. (cf. Rohdenburg 2006a). 100%
58
80% 208 60%
40%
20%
0%
0 Early Modern English
Late Modern English
Late 20th century
Figure 53.16: Diachronic development of plural marking with the measure noun fathom (N = 405) (based on Rohdenburg 2006a: 113)
862
VI Late Modern English
4 The adverb 4.1 Temporal adverbs The temporal adverb thrice has largely been superseded by its analytic variant three times after the LModE period. (29)
A woman should marry, – ; once, twice, and thrice if necessary. (1873 Trollope, The Eustace Diamonds)
(30)
At half-past seven Lord Fawn was brought into the room by his sister, and Andy Gowran, rising from his chair, three times ducked his head. (1873 Trollope, The Eustace Diamonds)
We also witness an incipient decline of twice to the benefit of two times. This substitution process is more advanced in AmE than in BrE (cf. Rohdenburg and Schlu¨ter 2009: 374).
4.2 Sentence adverbs The rapid expansion of evaluative sentence adverbs like sadly, oddly, curiously, surprisingly, plus enough is a relatively recent development (cf. Swan 1991: 418–419), which is particularly pronounced in BrE rather than in AmE. Swan sets up four semantic-pragmatic classes of sentence adverbs: truth intensifiers (e.g. surely), epitheticals (e.g. wisely), evaluative ADJ (e.g. miraculously) and speech act ADV (e.g. frankly). She observes that the overall number of sentence adverbs has been increasing since Old English, with truth intensifiers leading in all periods of English except Present-day English. As late as the Modern English period, a concomitant semantic diversification led to the establishment of speech act adverbs and a marked increase of evaluative sentence adverbs. This spread of evaluative sentence adverbs goes hand in hand with their greater positional freedom. What is more, BrE in the 1960s was far more advanced in dispensing with additional post-modification by enough (curiously enough, oddly enough), which is able to disambiguate the function of the sentence adverbial. By the 1990s, AmE catches up with the homeland variety as regards the frequency of using evaluation sentence adverbials without post-modification (cf. Rohdenburg and Schlu¨ter 2009: 380).
4.3 Variable adverb marking Another change concerning adverbs is the variable marking by the -ly suffix, in e.g. vast vs. vastly, great vs. greatly, fine vs. finely. While zero marking was the historically older form, explicit adverb marking generally increases throughout the history of English (cf. Tagliamonte 2009: 498). The speed of this development in Late Modern English is subject to rhythmic requirements: the adoption of the -ly suffix is generally delayed in attributive as opposed to predicative positions, presumably because in the former position -ly can act as prosodic buffer avoiding stress clashes (cf. Rohdenburg 2004: 356–357). Interestingly, a reversal of this trend towards more explicit adverb marking has recently been observable in both BrE and AmE, with the former spearheading this change.
53 Late Modern English: Morphology
5 The adjective The LModE period has often been described as a time when the standardization process promoted marked changes in the language system. It is also portrayed as a period with immense fluctuation. For instance, Poldauf (1948: 240) has stressed that 18th century rules governing comparative formation were less consistent than those operative in the 20th century. Bauer (1994) suggests that the system of comparative formation has become more systematic in the 20th century: The change in the course of this century appears to have been only incidentally an increase in the use of periphrastic comparison. Rather, the change has been a regularization of a confused situation, so that it is becoming more predictable which form of comparison must be used (Bauer 1994: 60; emphasis mine).
While it has often been assumed that English develops from a synthetic to a predominantly analytic language, this typological shift has also been claimed to hold for comparatives by Barber (1964: 131), Fries (1999: 96) and Quirk et al. (1985: 462n.) to name only a few. However, this generalization proves too coarse on two grounds: – it overlooks important changes within the system of comparison; – it holds no explanatory potential for uncovering the mechanisms determining comparative alternation and morphsyntactic change. The distribution of comparative variants in Late Modern English reveals an emergent functionally-motivated division of labor, with the analytic form becoming the domain of cognitively complex environments (e.g. if the adjective is accompanied by a prepositional or infinitival complement or if it is poly- rather than monomorphemic, etc.), while the synthetic form increases its range of applications in easier-to-process environments (cf. Mondorf 2009: 117–170). In order to trace this development, a look at the origin of analytic forms is in order. While some believe that the analytic variant emerged during the ME period as a result of French influence, Faiß (1977: 167) regards the more-comparative as a native English phenomenon, triggered by a strive for explicitness. We can adduce two types of evidence providing support for a system-internal emergence of analytic comparatives: (a) contrastive evidence from Germanic languages, and (b) Old English sources. The first explanation draws on the observation that the analytic comparative has developed independently in other Germanic languages than English as well (cf. e.g. Knu¨pfer 1921: 368–369; Nordberg 1985 on Swedish; Kyto¨ and Romaine 1997: 345 on Danish and Dutch; Mondorf 2009: 189 n. on individual cases in German). The second explanation adduces data from Old English (cf. Mitchell 1985: 84–85; Gonza´lez-Dı´az 2008: 28). Gonza´lez-Dı´az traces the emergence of analytic comparatives to the combination of participial forms with intensifying adverbs, occurring at least as early as the 9th century, as demonstrated in (31) and (32). (31)
þæt hi syn sylfe ma gode þonne oðre men ‘that they themselves are more good than other men’
863
864
VI Late Modern English (32)
& swyþe smælon leafon swylce heo ma fexede gesewen sy ‘and very small leaves in such a manner that she may appear to be more hairy’ (cf. Gonza´lez-Dı´az 2008: 19–20)
As regards the situation in Late Modern English, the pioneering study by Kyto¨ and Romaine (1997) shows that while 18th century authors still select the synthetic variant in 65% of all instances of willing, e.g. (33) (33)
the King was willinger to comply with any thing than this (18th century Defoe, Memoirs of a Cavalier 237)
the analytic variant is used almost exclusively (98%) in the 19th century. Conversely, disyllabic adjectives ending in often follow the reverse pattern by exhibiting a tendency from analytic > synthetic (cf. Kyto¨ and Romaine 1997: 343–344). The division of labor between the -er comparative and its more variant can be portrayed for 16 adjective groups that differ in length and suffix/final segment (cf. Mondorf 2009: 211–248). Lumping together all group results yields overall figures that show a decline in the analytic comparative. A differentiation by formal type reveals that this procedure conceals divergent developments. Only monosyllabic adjectives and disyllables in have decreased their share of analytic forms over time. 14 other groups (ending in -ful, -some, -ing, -ent, etc.) have either always formed the comparative analytically or have increased their share of analytic forms to up to 100%. – Only monosyllabic adjectives and disyllables ending in -y adhere to the pattern which predicts an incremental use of the synthetic variant at the expense of the analytic form, i.e. analytic > synthetic. – For the vast majority of adjective groups the trend is synthetic > analytic. – Many of those groups which allowed variation in past centuries are now knock-out contexts for the synthetic comparative. The system of comparison in Late Modern English develops a remarkably clear pattern of reorganization. Comparative alternation becomes more systematic by exploiting the options offered by the availability of morphosyntactic variants. As regards processing requirements it makes the best possible use of each variant: cognitively complex environments (such as adjectives accompanied by a complement) consistently increase their share of the analytic comparative to such an extent that the analytic form can even become obligatory. By contrast, easy-to-process environments can afford to use the synthetic comparative. What is at stake in the diachronic development of comparative alternation is an economically-motivated division of labor, in which each variant assumes the task it is best suited to. This division of labor emerges in the LModE period, around the 18th century (cf. Figure 53.16 below). The benefit of creating a typologically consistent, uniform synthetic or analytic language system is outweighed by forces following functional motivations. The older synthetic form can stand its ground (or even reconquer formerly lost domains) in easy-to-process environments, while the analytic form prevails in cognitively-demanding contexts. This emergent division of labor also accounts for the enhanced processing requirements associated with argument complexity. Both the synthetic and the analytic
53 Late Modern English: Morphology
865
variant are quite freely used in the 17th century in the presence of infinitival (34) and prepositional (35) complements. (34)
since, I truely speake, I thinke none worthyer to bee truely loued (except my Lord) then I imagine him (1621 Lady Mary Wroth, The Countesse of Mountgomeries Urania)
(35)
thy mistris is more worthie of honour, then our whole countrey (1635 Breton, A Mad VVorld my masters)
By contrast, the 19th century already offers a range of examples illustrating the emergent division of labor: in the absence of complements, the synthetic comparative variant tends to be favored (36), while in the presence of complements the analytic variant has been preferred (37). (36)
It was not perhaps that he himself imagined that Tregear as a member of Parliament would be worthier, but that he fancied that such would be the Duke’s feelings. (1880 Trollope, The Duke’s Children)
(37)
I only wish I were more worthy of him. (1818 Austen, Northanger Abbey)
Figure 53.17 provides the distribution of synthetic vs. analytic comparatives in nonattributive position in the presence or absence of complements in British prose data. The line with squares indicates the use of the more-variant in the presence of a prepositional or infinitival complement. The line with triangles depicts uses of the analytic comparative in the absence of such complements. 100%
Analytic Comparative %
112 80%
+ complement − complement
125
60% 48 76
211 612
40% 200
414 20%
0% 1500–1700
1700 –1800
1800 –1900
1960 –1993
full, keen, lucky, pleasant, proud, ready, sure, worthy
Figure 53.17: Diachronic development of analytic comparatives +/-complement from Early Modern to Present-day English in BrE (NSynthetic + Analytic = 5987) (based on Mondorf 2009: 251)
866
VI Late Modern English The 19th and 20th centuries are crucial for the emergent division of labor. The use of analytic forms in environments that require an increased processing load must have emerged around the 18th century. It looks as if the English comparative system has only fairly recently settled to some extent the old conflict in which two variants compete for their own territory. The outcome – i.e. that the more-variant is required with long, morphologically complex, lexically complex and less frequent words – can be accounted for, if we assume that analytic structures are more explicit and hence easier to process (cf. Mondorf 2009: 9). Finally, we ought to address the long-standing issue of the direction of change. A frequently mentioned pattern concerning the morphology-syntax interface in language change assumes a cyclic movement of alternating synthetic and analytic structures (cf. e.g. Givo´n 1979). What we are observing in comparative alternation is divergent developments even within one and the same structure, thereby exploiting the respective processing advantages of synthetic and analytic forms.
6 Summary Despite the fact that the short time-span of merely two centuries is likely to witness morphological developments that either had their incipient stages before the LModE period or that do not reach completion within this period, the existence of far-reaching morphological changes that come about in such a short time-period prove remarkable and instructive. Rather than rendering the picture of a “confused situation” (Poldauf 1948: 240), the high degree of variability in combination with the observed trajectories of change described in the present chapter can be shown to conceal systematic developments which are well in line with functionally motivated principles. Acknowledgments: I wish to thank Gu¨nther Lampert and Matthias Eitelmann for very helpful comments on an earlier draft of this paper, as well as Anna-Lena Krebs for retrieving the data on acquaintance.
7 References Auer, Anita. 2006. Precept and practice: The influence of prescriptivism on the English subjunctive. In: Dalton-Puffer et al. (eds.), 33–54. Auer, Anita. 2007. Lest the situation deteriorates: A study of lest as trigger of the inflectional subjunctive. In: Miriam A. Locher and Ju¨rg Stra¨ssler (eds.), Standards and Norms in the English Language, 149–173. Berlin/New York: Mouton de Gruyter. Barber, Charles L. 1964. Linguistic Change in Present-Day English. London: Andre´ Deutsch. Bauer, Laurie. 1994. Watching English Change: An Introduction to the Study of Linguistic Change in Standard Englishes in the Twentieth Century. London: Longman. Crawford, William J. 2009. The mandative subjunctive. In: Rohdenburg and Schlu¨ter (eds.), 257–276. Dalton-Puffer, Christiane, Dieter Kastovsky, Nikolaus Ritt, and Herbert Schendl (eds.). 2006. Syntax, Style and Grammatical Norms: English from 1500–2000. Frankfurt: Peter Lang. Davies, Mark. 2008–. The Corpus of Contemporary American English: 425 million words 1990–present. Available online at http://corpus.edu/coca Dekeyser, Xavier. 1975. Number and Case Relations in 19th Century British English: A Comparative Study of Grammar and Usage. Antwerpen: De Nederlandsche Boekhandel. Denison, David. 1998. Syntax. In: Suzanne Romaine (ed.), The Cambridge History of the English Language. Vol. IV. 1776–1997, 92–329. Cambridge: Cambridge University Press. Faiß, Klaus. 1977. Aspekte der englischen Sprachgeschichte. Tu¨bingen: Narr.
53 Late Modern English: Morphology Fanego, Teresa. 1996. On the historical development of English retrospective verbs. Neuphilologische Mitteilungen 97: 71–79. Fischer, Olga C. M. 1997. The grammaticalisation of infinitival to in English compared with German and Dutch. In: Raymond Hickey and Stanisław Puppel (eds.), Language History and Linguistic Modelling. A Festschrift for Jacek Fisiak on his 60th Birthday. Vol. 1: Language, 265–280. Berlin/New York: Mouton de Gruyter. Fowler, Henry W. 1925. Fused participle. Society for Pure English Tract 22: 43–47. Fries, Charles C. 1999. Postnominal modifiers in the English noun phrase. In: Peter Collins and David Lee (eds.), The Clause in English, 93–110. Amsterdam/Philadelphia: John Benjamins. Givo´n, Talmy. 1979. From discourse to syntax: Grammar as a processing strategy. In: Talmy Givo´n (ed.), Syntax and Semantics 12: Discourse and Syntax, 81–109. New York: Academic Press. Go¨rlach, Manfred. 1999. English in Nineteenth-century England: An Introduction. Cambridge: Cambridge University Press. Go¨rlach, Manfred. 2001. Eighteenth-century English. Heidelberg: Winter. Gonza´lez-Dı´az, Victorina. 2008. English Adjective Comparison: A Historical Perspective. Amsterdam/Philadelphia: John Benjamins. Heyvaert, Liesbet, Hella Rogiers, and Nadine Vermeylen. 2005. Pronominal determiners in gerundive nominalization: A case study. English Studies 86(1): 71–88. Hudson, Richard. 2003. Gerunds without phrase structure. Natural Language and Linguistic Theory 21: 579–615. Hundt, Marianne. 2009. Colonial lag, colonial innovation or simply language change? In: Rohdenburg and Schlu¨ter (eds.), 13–37. Israel, Michael. 1996. The way constructions grow. In: Adele Goldberg (ed.), Conceptual Structure, Discourse and Language, 217–230. Stanford: CSLI. Jespersen, Otto. 1926. On some disputed points in English grammar. On ING. Society for Pure English Tract 25: 147–172. Jespersen, Otto. 1961 [1928]. A Modern English Grammar on Historical Principles. Part III: Syntax (Second Volume). Northampton: J. Dickens & Co. Kjellmer, Go¨ran. 1980. Accustomed to swim: Accustomed to swimming. On verbal forms after TO. In: Jens Allwood and Magnus Ljung (eds.), ALVAR: A Linguistically Varied Assortment of Readings: Studies Presented to Alvar Ellegard on the Occasion of his 60th Birthday, 75–99. Stockholm: Almqvist and Wiksell. Kjellmer, Go¨ran. 2009. The revived subjunctive. In: Rohdenburg and Schlu¨ter (eds.), 246–256. Knu¨pfer, Hans. 1921. Die Anfa¨nge der periphrastischen Komparation im Englischen. Englische Studien 55: 321–389. Kyto¨, Merja. 1993. Third-person present singular verb inflection in Early British English and American English. Language Variation and Change 5: 113–139. Kyto¨, Merja and Suzanne Romaine. 1997. Competing forms of adjective comparison in Modern English: What could be more quicker and easier and more effective? In: Terttu Nevalainen and Tarkka Leena Kahlas (eds.), To Explain the Present: Studies in the Changing English Language in Honour of Matti Rissanen, 329–352. Helsinki: Socie´te´ Ne´ophilologique. Lass, Roger. 1999. Phonology and morphology. In: Roger Lass (ed.), The Cambridge History of the English Language. Vol. III. 1476–1776, 56–186. Cambridge: Cambridge University Press. Levin, Magnus. 2001. Agreement with Collective Nouns in English. Lund: Lund University Press. Levin, Magnus. 2006. Collective nouns and language change. English Language and Linguistics 10(2): 321–343. Liedtke, Ernst. 1910. Die numerale Auffassung der Kollektiva im Verlaufe der englischen Sprachgeschichte. Ph.D. Dissertation, Albertus-Universita¨t zu Ko¨nigsberg. Lyne, Susanna. 2006. The form of the pronoun preceding the verbal gerund: Possessive or objective? ICAME Journal 30: 37–53. Lyne, Susanna. 2007. Subjects of verbal gerunds in Late Modern English. Paper presented at the 28th ICAME conference in Stratford-on-Avon, UK, May 23–27, 2007.
867
868
VI Late Modern English Mair, Christian. 1993. A crosslinguistic functional constraint on believe-type raising in English and selected other European languages. In: Jacek Fisiak (ed.), Papers and Studies in Contrastive Linguistics. Vol. 28, 5–19. Poznan´: Adam Mickiewicz University. Mair, Christian. 2006. Nonfinite complement clauses in the nineteenth century: The case of remember. In: Merja Kyto¨, Mats Ryde´n, and Erik Smitterberg (eds.), Nineteenth-century English. Stability and Change, 215–228. Cambridge: Cambridge University Press. Marckwardt, Albert H. 1958. American English. London: Oxford University Press. Milroy, James and Lesley Milroy (eds.). 1993. Real English: The Grammar of English Dialects in the British Isles. London: Longman. Mitchell, Bruce. 1985. Old English Syntax. 2 vols. Oxford: Clarendon Press. Mondorf, Britta. 2009. More Support for More-Support: The Role of Processing Constraints on the Choice between Synthetic and Analytic Comparative Forms. Amsterdam/Philadelphia: John Benjamins. Mondorf, Britta. 2010. Genre-effects in the replacement of reflexives by particles. In: Heidrun Dorgeloh and Anja Wanner (eds.), Approaches to Syntactic Variation and Genre, 219–246. Berlin/ New York: Mouton de Gruyter. Mondorf, Britta. 2011. Variation and change in English resultative constructions. Language Variation and Change 22: 397–421. Nordberg, Bengt. 1985. Vad ha¨der med adjektivets a¨ndelsekomparation? In: Bengt Nordberg (ed.), Det Ma˚ngskiftande Spra˚ket. Om Variation i Nusvenskan, 88–103. Lund: Liber Fo¨rlag. ¨ vergaard, Gerd. 1995. The Mandative Subjunctive in American and British English in the 20th O Century. Stockholm: Almqvist & Wiksell. Peters, Pam. 2004. The Cambridge Guide to English Usage. Cambridge: Cambridge University Press. Poldauf, Ivan. 1948. On the History of Some Poblems of English Grammar before 1800. Prague: Na´kladem Filosoficke´ Fakulty University Karlovy. Quirk, Randolph, Sidney Greenbaum, Geoffrey Leech, and Jan Svartvik. 1985. A Comprehensive Grammar of the English Language. London: Longman. Rohdenburg, Gu¨nter. 2003a. Cognitive complexity and horror aequi as factors determining the use of interrogative clause linkers in English. In: Rohdenburg and Mondorf (eds.), 205–250. Rohdenburg, Gu¨nter. 2003b. Aspects of grammatical iconicity in English. In: Wolfgang Mu¨ller and Olga Fischer (eds.), From Sign to Signing, 263–285. Amsterdam/Philadelphia: John Benjamins. Rohdenburg, Gu¨nter. 2004. Comparing grammatical variation phenomena in non-standard English and Low German dialects from a typological perspective. In: Bernd Kortmann (ed.), Dialectology Meets Typology: Dialect Grammar from a Cross-linguistic Perspective, 335–366. Berlin/New York: Mouton de Gruyter. Rohdenburg, Gu¨nter. 2006a. Variable plural marking with measure nouns in non-standard English and Low German dialects. NOWELE: Studies in North-Western European Language Evolution 48: 111–130. Rohdenburg, Gu¨nter. 2006b. The role of functional constraints in the evolution of the English complementation system. In: Dalton-Puffer et al. (eds.), 143–166. Rohdenburg, Gu¨nter. 2009a. Reflexive structures. In: Rohdenburg and Schlu¨ter (eds.), 166–181. Rohdenburg, Gu¨nter. 2009b. Grammatical divergence between British and American English in the 19th and early 20th centuries. In: Ingrid Tieken-Boon von Ostade and Wim van der Wurff (eds.), Current Issues in Late Modern English, 287–315. Bern: Lang. Rohdenburg, Gu¨nter and Britta Mondorf (eds.). 2003. Determinants of Grammatical Variation in English. Berlin/New York: Mouton de Gruyter. Rohdenburg, Gu¨nter and Julia Schlu¨ter. 2009. New departures. In: Rohdenburg and Schlu¨ter (eds.), 364–423. Rohdenburg, Gu¨nter and Julia Schlu¨ter (eds.). 2009. One Language, Two Grammars: Grammatical Differences between British English and American English. Cambridge: Cambridge University Press.
54 Late Modern English: Syntax
869
Schlu¨ter, Julia. 2009a. Phonology and grammar. In: Rohdenburg and Schlu¨ter (eds.), 108–129. Schlu¨ter, Julia. 2009b. The conditional subjunctive. In: Rohdenburg and Schlu¨ter (eds.), 277–305. Swan, Toril. 1991. Adverbial shifts: Evidence from Norwegian and English. In: Dieter Kastovsky (ed.), Historical English Syntax, 409–438. Berlin/New York: Mouton de Gruyter. Tagliamonte, Sali. 2009. Historical change in synchronic perspective: The legacy of British dialects. In: Ans van Kemenade and Bettelou Los (eds.), The Handbook of the History of English, 477–506. Oxford: Blackwell. Tieken-Boon van Ostade, Ingrid. 2009. An Introduction to Late Modern English. Edinburgh: Edinburgh University Press. Vosberg, Uwe. 2003. The role of extractions and horror aequi in the evolution of -ing complements in Modern English. In: Rohdenburg and Mondorf (eds.), 305–327. Vosberg, Uwe. 2009. Non-finite complements. In: Rohdenburg and Schlu¨ter (eds.), 212–227. Warner, Anthony R. 1982. Complementation in Middle English and the Methodology of Historical Syntax. London: Croom Helm.
Britta Mondorf, Mainz (Germany)
54 Late Modern English: Syntax 1 2 3 4 5
Introduction Categorical innovations of the Late Modern English period Statistical and regulatory changes Summary References
Abstract The Late Modern English period provides an essential link between the syntactic innovations of Early Modern English and the established system of Present-day English. This chapter reviews a number of major syntactic developments taking place in the period, involving both categorical and statistical changes. Among the former, we discuss two important innovations in the domain of voice: the rise of the progressive passive, with its implications for the symmetry of the auxiliary system, and the emergence and consolidation of the get-passive. The 18th and 19th centuries also see the completion and/or regulation of long-term tendencies in various areas of syntax, such as the verb phrase (the consolidation of the progressive, the decline of the be-perfect and the regulation of periphrastic do), and subordination (changes in the complementation system, in particular the replacement of the to-infinitive by -ing complements, and in relative clauses, with the regulation of the distribution of the different relativizers).
1 Introduction The Late Modern English period has received much less scholarly attention than earlier stages in the history of English, partly because of its closeness to the present day and its apparent similarity to the contemporary language. This neglect has been particularly noticeable in the case of syntax. It is a well-known fact that the most substantive Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 869–887
870
VI Late Modern English syntactic changes in the history of English had already taken place when our period opened, the 18th and 19th centuries representing mainly a transitional stage between the categorical innovations of Late Middle English and, especially, Early Modern English and the “established” system of Present-day English. It is not at all surprising, therefore, that the developments which occur in the Late Modern English period are confined, with a few notable exceptions discussed below, to changes concerning the regulation of variants introduced in previous periods, with certain patterns or constructiontypes becoming more frequent than others, and to the consolidation of processes which had been in progress for some time. We believe, however, that the importance of Late Modern English syntax should not be underestimated, and that we should look at the 18th and 19th centuries in our search for the definitive link to Present-day English syntactic usage. In what follows we will first consider the most important categorical innovations of the period (Section 2), and then move on to the discussion of individual selected areas of syntax affected by statistical and regulatory changes between 1700 and the early 20th century (Section 3). The analysis of these changes has undoubtedly been facilitated by the wealth of language material representing the 18th and 19th centuries in comparison to earlier stages, as well as by the availability in recent years of computerized corpora covering our period, among others, A Representative Corpus of Historical English Registers (ARCHER) (Biber and Finegan 1990–93/2002/2007/ 2010), A Corpus of Nineteenth-century English (CONCE) (Kyto¨ and Rudanko forthc.), The Corpus of late Modern English Prose (Denison et al. 1994), The Century of Prose Corpus (COPC) (Milic 1995), The Corpus of Late Modern British and American English Prose (Fanego 2007a), and databases such as Chadwyck-Healey’s Eighteenthcentury Fiction (ECF) (Hawley et al. 1996) and Nineteenth-century Fiction (NCF) (Karlin and Keymer 2000). Our account of Late Modern English syntax is, of necessity, incomplete. Other relevant changes, such as the expansion of multi-word verbs (e.g. give over, put up with, take a look), the development of the prop-word one or of the so-called emerging modals (e.g. have to, have got to, want to; cf. Krug 2000), the last steps in the fixing of SVO order, changes in the syntax of the noun phrase, and the influence of prescriptivism on the avoidance of preposition stranding and the split infinitive, among others, have been left out of this overview article due to limitations of space. For information on these and other Late Modern English syntactic changes, the reader is particularly referred to Denison’s (1998) comprehensive and authoritative chapter in volume IV of the Cambridge History of the English Language (Romaine [ed.] 1998).
2 Categorical innovations of the Late Modern English period It is the domain of voice that witnessed the most important categorical changes taking place in the course of Late Modern English: on the one hand, the emergence of the progressive passive and, on the other, the grammaticalization of the get-passive construction.
2.1 The rise of the progressive passive One of the few grammatical innovations of the Late Modern English period is the development of the progressive passive construction (e.g. The patient is being
54 Late Modern English: Syntax examined). This pattern emerged towards the end of the 18th century. Before that time, the progressive was either avoided, as in (1) – in Present-day English one would say was being dragged – or the so-called passival (i.e. an active progressive with passive meaning) was used, as in (2) (both examples from Denison 1998: 148, 151). (1)
he found that the coach had sunk greatly on one side, though it was still dragged forward by the horses; (1838–39 Dickens, Nicholas Nickleby v.52)
(2)
But are there six labourers’ sons educating in the universities at this moment? (1850 Kingsley, Alton Locke xiii.138)
Early instances of the progressive passive construction are given in (3) and (4) below. Example (4) illustrates the coexistence of the old passival and the new progressive passive in the early 19th century. (3)
like a fellow whose uttermost upper grinder is being torn out by the roots by a mutton-fisted barber. (1795 Southey I.249.24; Traugott 1972: 178)
(4)
While the goats are being milked, and such other refreshments are preparing for us as the place affords. (1829 Landor, Imag. Conv., Odysseus, etc.; Denison 1993: 428)
The progressive passive pattern was prescriptively condemned. Nineteenth-century purists declared it to be “not English”, “an outrage upon English”, “a monstrosity”, a “corruption of language”, etc. (Bailey 1996: 222–223). Nonetheless, the construction firmly became part of the English language in the course of the period, perhaps because, as Denison (1998: 151) notes, “the adoption of the progressive passive makes the English auxiliary system much more symmetrical”.
2.2 The grammaticalization of the get-passive construction Another categorical change affecting the domain of the passive in the course of the Late Modern English period is the emergence and consolidation of the get-passive construction (e.g. He got fired from his first job). According to Strang (1970: 151) and Givo´n and Yang (1994: 131), it is not until the second half of the 18th century that unequivocal examples of the get-passive are found. Some earlier instances do, however, occur (cf. 5 and 6). Notice that other early candidates of the construction (cf. 10 below) are somewhat dubious and could be interpreted as involving a predicative structure rather than a true passive construction (cf. Strang 1970: 150–151; Denison 1993: 420). (5)
I am resolv’d to get introduced to Mrs. Annabella; (1693 Powell, A Very Good Wife II.i.p.10; Denison 1998: 320; Gronemeyer 1999: 29)
(6)
So you may not save your life, but get rewarded for your roguery (1731 Fielding, Letter Writers II.ix.20; Denison 1993: 420)
871
872
VI Late Modern English The early 19th-century instances in (7– 9), all from Denison (1993: 434, 436), illustrate the occurrence of the get-passive in combination with auxiliary verbs and with the progressive, which testifies to the rapid expansion of the construction in our period. (7)
I shall get plentifully bespattered with abuse (1819 Southey, Letters; OED, s.v. bespatter)
(8)
Her siren finery has got all besmutched (1832 Carlyle in Fraser’s Mag. V.258; OED, s.v. besmutch v.)
(9)
My stomach is now getting confirmed, and I have great hopes the bout is over (1819 Scott, Let. in Lockhart [1837] IV viii 253; OED, s.v. set-to, def. 2b)
It is not until the 20th century, however, that the get-passive becomes firmly established. Hundt’s (2001: 85, Table 3) raw figures for the combination get + past participle in the ARCHER corpus confirm the gradual increase of the construction through time: 11 examples in the 18th century, 26 instances in the 19th century, and 75 occurrences in the 20th century. Of particular interest in the history of the get-passive is its controversial origin. On the basis of the evidence from a corpus of data from the mid-14th to the mid-20th century, Givo´n and Yang (1994: 144–145) suggest that the construction originates in the causative use of the verb get through a process of reflexivization and de-transitivization along the following lines: She got him to be admitted (causative structure with be-passive complement) > She got herself to be admitted (causative-reflexive construction with bepassive) > She got to be admitted (intransitive-inchoative structure) > She got admitted (get-passive construction). Givo´n and Yang acknowledge, however, that other related structures, among them the intransitive-locative construction (e.g. She got into the house) and the inchoative-adjectival get-construction (e.g. She got anxious), may also have played a role in the development via analogy. Givo´n and Yang’s account has been called into question in a number of recent studies. In her investigation of the relationship between frequency and the grammaticalization of the get-passive, Hundt (2001: 64–67) concludes that the importance attached by Givo´n and Yang to reflexive constructions in the process of change may have been overestimated. On the contrary, the evidence from ARCHER seems to indicate that causative passives without be may have played a more crucial role in the grammaticalization of the get-passive than assumed by Givo´n and Yang (Hundt 2001: 67–68). An alternative diachronic pathway of development has been proposed by Gronemeyer (1999) and, especially, Fleisher (2006), who maintain that it is the inchoative use of get (e.g. She got anxious), rather than its causative value that lies at the root of the get-passive. In Gronemeyer’s (1999: 29) view, “the get-passive evolved out of the inchoative construction when the matrix subject is reanalysed as controlling the implicit internal argument of the participle, rather than the implicit external one as in the inchoative”. Fleisher (2006) also defends the inchoative-to-passive pathway identified by Gronemeyer, but takes a novel perspective on the mechanisms responsible for the change. In his view, structural ambiguity is not the only factor leading to reanalysis, but semantic and pragmatic forces (e.g. perfective aspect in the case
54 Late Modern English: Syntax of the get-passive) are also among the critical motivations for syntactic change. He maintains that Givo´n and Yang’s causative source hypothesis is highly problematic syntactically, semantically, and thematically (Fleisher 2006: 239–245) and, using the drama and prose sections of the Literature Online corpus (LION) (Chadwyck-Healey 1996–2011), he identifies two stages in the development of passive get from inchoative get: (i) up to 1760, when the past participles occurring with get are interpretable as adjectives and can therefore yield an inchoative reading, as in example (10) below; and (ii) after 1760, when such a restriction disappears and the class of participles entering the construction is expanded considerably, as shown in (11) (Fleisher 2006: 227, 230–232). (10)
A certain Spanish pretending Alchymist … got acquainted with foure rich Spanish merchants … (1652 Gaule, Magastrom 361; OED, s.v. get v., def. 34b; Denison 1993: 419; Fleisher 2006: 227)
(11)
from thence you got expell’d for robbing the poors’ box (1778 Foote, A Trip to Calais; Fleisher 2006: 231)
3 Statistical and regulatory changes In addition to the categorical innovations discussed so far, the Late Modern English period witnessed the completion of a wide variety of changes which had started in earlier stages and the regulation of variants in several syntactic domains. This section offers a representative selection of such statistical and regulatory changes in the verb phrase (progressive, perfect, and auxiliary do) and in subordination (complementation and relative clauses).
3.1 The progressive The progressive construction has been part of the English language since Old English times (Traugott 1992: 187; Denison 1993: 371) and became an established pattern in Early Modern English (Denison 1998: 130; see Seoane, Chapter 39). Its core use is to express an ongoing event. Noteworthy during the 19th century is a marked increase in the use of the progressive, described in various sources, among them Dennis (1940), Strang (1982), Arnaud (1983), Denison (1998), Hundt (2004), Smitterberg (2005), and Nu´n˜ez-Pertejo (2007). Strang (1982: 442) observes an increase of the progressive in narrative prose during the period which was mostly observable in main clauses. She also points to its expanded use with different types of main verbs and in combination with perfective, passive, and modal auxiliaries (Strang 1982: 452–453). Using data from CONCE, Smitterberg (2005: 57–58) makes a case for the progressive becoming integrated in the English language during the 19th century, where “integration” is understood to cover grammaticalization, obligatorification, the use of the progressive in combination with other verbs, and the extended range of uses of the progressive in different situations. The increase of the use of the progressive in a number of genre categories is shown in Figure 54.1.
873
874
VI Late Modern English 500
M-coefficient
400 Debates Drama Fiction History Letters Science Trials
300
200
100
0 1
2 Period
3
Figure 54.1: The increased use of the progressive in 19th-century English. The three periods are delimited as follows: period 1: 1800−1830; period 2: 1850−1870; period 3: 1870−1900 (from Smitterberg 2005: 66)
Using the Mosse´-coefficient (M-coefficient), which calculates the number of occurrences per 100,000 words, this figure shows an increase in the use of the progressive in all but one genre over the three periods. The increase in the letters genre and in the speech-based genre of drama is especially notable. Smitterberg’s (2005: 88) conclusion is that “the progressive increased considerably in frequency over the 19th century, though not quite so much as other scholars have reported”. We can compare the number of instances of the progressive in Smitterberg’s data with the figure found by Nu´n˜ez-Pertejo (2007: 363), who studied the progressive in the 18th century using British English data from the COPC and ARCHER. She reports a much lower overall frequency of the construction of 77.6 per 100,000 words in the relevant subsections of the combined corpora, as shown in Table 54.1. (For ease of comparison with the data supplied by other authors discussed in this chapter, we have normalized the frequencies of Nu´n˜ez-Pertejo’s data to instances per 100,000 words.) This figure compares with her finding (Nu´n˜ez-Pertejo 2004) of 35.6 instances per 100,000 words in the subperiod 1640–1710 of the Early Modern English part of the Helsinki Corpus (Rissanen et al. 1991). Table 54.1: The frequency of the progressive in COPC and ARCHER (from Nu´n˜ez-Pertejo (2007: 363) (NF = normalized frequency per 100,000 words) COPC Subperiod
#
ARCHER NF
Subperiod
1680–1699 46 63.5 1700–1749 125 61.3 1700–1749 1750–1780 114 60.1 1750–1780 Subtotal 285 61.9 Subtotal Total in the combined corpora (285+347): 632; NF: 77.6
# 160 187 347
NF 91.8 107 99.7
54 Late Modern English: Syntax
875
Work by Hundt (2004) has shown that the increase in the use of the progressive continued during the 20th century. Mair and Leech (2006: 323; based on Mair and Hundt 1995) studied the use of the progressive in written Present-day (British and American) English during the period 1961–91, finding a similar trend. They compared the frequency of the progressive in LOB (Leech and Johanson 1978) with that in F-LOB (Mair 1998) and its frequency in Brown (Francis and Kucˇera 1979 [1964]) with that in Frown (Mair 1999). The results appear in Table 54.2. Table 54.2: Progressive forms in the press sections (A–C) of the four reference corpora (significances: LOB p < 0.01, Brown p < 0.05; LOB-Brown and F-LOB-Frown p > 0.05) (from Mair and Leech 2006: 323)
British English (LOB/F-LOB) American English (Brown/Frown)
1961
1991/92
Difference (percentage of 1961)
606 593
716 663
+18.2% +11.8%
These data show that the frequency of use of the progressive from the 1960s to the 1990s increased in written British and American English. All the sources mentioned so far are based on written data. Aarts et al. (2010) use the Diachronic Corpus of Present-Day Spoken English (DCPSE) (Aarts and Wallis 2006), which contains 400,000 words from the London-Lund Corpus (LLC; 1970s) (Greenbaum and Svartvik 1990) and 400,000 words from the British component of the International Corpus of English (ICE-GB; 1990s) (Greenbaum and Nelson 1998), to study the use of the progressive construction in spoken English. The results are further evidence of an increase in the use of the progressive during the 20th century. Naturally, some care in interpreting the data from the various sources is called for, given possible differences in the composition of the corpora (text types) and differences in how progressive constructions are counted (e.g. is be going to excluded?). Nevertheless, it seems unmistakeably the case that the use of the progressive has increased steadily over the last few centuries. What could be the explanation for an increased use of the progressive? It appears that an expansion of the use of particular meanings has played a role. Nesselhauf (2007) has found that the “progressive futurate” (e.g. I’m leaving tomorrow) tripled in use between 1750 and 1990, while others, among them Wright (1994, 1995), Smith (2005), and Smitterberg (2005), have suggested that the so-called “interpretive”, “explanatory”, or “modal” use of the progressive, as in (12) below, which experiences a marked increase in the 19th century (cf. Kranich 2009), has contributed to the increase in frequency of the progressive construction, particularly of the present progressive, in British English. (12)
If the Government claim they have reduced taxes, they are hoodwinking you.
Here the clause in the progressive is said to furnish an interpretation or explanation of the situation mentioned in the subordinate clause. This may well be the case, though some linguists have cautioned against distinguishing different functions for the progressive. Thus, Visser is said by Denison (1998: 145) to be “taking a ruthless line against those who find a multiplicity of functions” for the progressive. In any case, Smitterberg
876
VI Late Modern English (2005) notes that his results regarding the use of the interpretive progressive during the 19th century are inconclusive as regards an increase in its use, due to the low incidence of the construction, though he does note that there is a trend for it to be used in speechbased genres. Mair and Hundt (1995: 111), writing about the 20th-century increased use of the progressive, deny that an expanded functional load plays a role, and claim that “[t]he increase is shown to be due to a growing tendency to use the progressive in cases where it has long competed with the simple form and not, as is sometimes alleged, to the establishment and spread of new forms and uses of the progressive”. The increase can be explained, they argue, either by pointing to the “colloquialization” of written English, or by observing that “the change consists in the fact that in cases in which the simple form can be used alongside the progressive, the latter tends to be chosen with increasing frequency – to the point that an originally marked or rare ‘progressive’ comes to constitute the statistical norm” (Hundt 1995: 118). They suggest that both explanations play a role, with colloquialization being the dominant one. On colloquialization pertaining to the progressive and phrasal verbs in the 19th century, see Smitterberg (2008). Further research on colloquialization is nevertheless needed both for Late Modern English and Present-day English.
3.2 The decline of be as perfect auxiliary One of the most conspicuous statistical changes taking place in our period concerns the variation between be and have as perfect auxiliaries with verbs of motion and mutation (e.g. go, come, grow, become, etc.), as shown in (13) and (14). (13)
A young man has gone to the happy hunting grounds. (1826 Cooper, Last of Mohicans [1831] 400; OED, s.v. hunting-ground)
(14)
You would be sorryish to hear, that poor Moll Cobb … is gone to her long home. (1793 Seward, Lett. [1811] III. 330; OED, s.v. sorryish a.)
The process of decline of be in such a context, which had already started in Late Old English and continued steadily in Middle English and Early Modern English (see Seoane, Chapter 39), was fairly well advanced when the Late Modern English period opened. Although the use of have with these verbs increased considerably from 1700 to 1800 (from 20% to 38% in Ryde´n and Brorstro¨m’s [1987] corpus of comedies and private letters; from 39% to 56% in Kyto¨’s [1997: 33] analysis of the multi-genre data in the ARCHER corpus), it was the 19th century that saw the most rapid changes in the encroachment of have upon the be-domain (85% of intransitives in ARCHER take have instead of be in the period 1800–1900; cf. Kyto¨ 1997: 33). The proportion of mutative verbs taking be as perfect auxiliary seems to have been reduced by half during the first decades of the 19th century, have almost completely taking over by the turn of the century (92% of the relevant cases in Ryde´n and Brorstro¨m’s data 1987: 198, 200). In addition to the attacks on the use of perfect be by 18th- and 19th-century grammarians (cf. Visser 1963–73: Section 1898; Ryde´n and Brorstro¨m 1987: 210), internal factors such as analogy (non-mutative verbs have always been more numerous than mutative ones), the neutralization of the is/has distinction under the clitic form ’s, and the avoidance of potential ambiguity (be is functionally more heavily overloaded than
54 Late Modern English: Syntax have) played a decisive role in the process of change (Traugott 1972: 145; Ryde´n and Brorstro¨m 1987: 23, 197, 287; Denison 1993: 366, 1998: 136). Typical contexts strongly favoring the choice of have in Ryde´n and Brorstro¨m’s data of Late Modern English informal written prose include, among others, the following (cf. also Kyto¨ 1997: 56–59): so-called action contexts (i.e. iteration and duration), unreality or uncertainty contexts (e.g. conditional and optative clauses, negative and questioned contexts), and the perfect infinitive. The early 18th-century examples in (15) and (16), both from Denison (1993: 368), illustrate the use of have with a mutative verb in an iterative context and a conditional clause, respectively. (15)
The letters have come so regularly of late that … (1714 Wentworth 383.3)
(16)
if he had not come up as he did he would have had a Feaver or convultions (1717 Verney I 397.22)
Additional factors conditioning the use of be and have with intransitive verbs in the period are (i) text type (e.g. journals show higher figures for the have auxiliary than other text types in the 18th-century ARCHER data; cf. Kyto¨ 1997: 44, Table 12); (ii) gender (women writers are more conservative than men writers in the Late Modern English subperiods in ARCHER; cf. Kyto¨ 1997: 50–51); and (iii) the nature of the main verb (for example, loan verbs attracted “the use of a rising [have] rather than a receding [be] form” (Kyto¨ 1997: 64), while the high-frequency items go and come retained the use of be much longer than other verbs; cf. Kyto¨ 1997: 67, Table 29).
3.3 The regulation of periphrastic do The 18th and 19th centuries also saw the last steps in the regulation of the uses of the so-called dummy auxiliary do. Periphrastic do appeared for the first time in the Late Middle English period, and gradually became more common in the following centuries (see Warner, Chapter 47). By 1700 its use was already very close to the modern one, with do being frequent, though not obligatory yet, in negative and interrogative constructions, as well as in affirmative declarative clauses, especially with an emphatic function. In non-emphatic affirmative constructions, however, the use of do was considered redundant and superfluous by the mid-18th century, a “vitious mode of speech” in Dr. Johnson’s (1979 [1755]: 8) words. In her detailed analysis of an 18th-century corpus of informative prose, epistolary prose, and dialogue in plays and novels, Tieken-Boon van Ostade (1987) shows that this use of do was indeed dying out by that time. Obviously enough, this “superfluous” do survived longer in poetry, given its usefulness as a line-filler. Consider in this connection example (17) below from Wordsworth. (17)
The hapless creature which did dwell / Erewhile within the dancing shell. (1827 Wordsworth, The Blind Highland Boy 193–194; Beal 2004: 73)
Tieken-Boon van Ostade’s (1987) data also show that non-periphrastic negation and question formation continued to occur in the early part of the Late Modern English period (an average of 24% and 5% of the relevant cases, respectively), though doless structures gradually declined through the 18th century. Factors such as style and
877
878
VI Late Modern English the author’s background influenced the use/non-use of periphrastic do (cf. Tieken-Boon van Ostade 2006: 261). In Tieken-Boon van Ostade’s (1987) view, the normative grammars of the time seem also to have played an important role in the regulation of the use of the auxiliary in these contexts, so that the process could be characterized, in Labovian terms, as a change from above. Robert Lowth (1979 [1775]: 41), for example, maintains that the auxiliary do “is of frequent and almost necessary use in interrogative and negative sentences” [emphasis added]. The old pattern of negation without do seems to have been particularly persistent with high-frequency verbs, such as know and doubt (I know not, I doubt not) (cf. Tieken-Boon van Ostade 1987: 128–129, 158–160, 174). The preservation of structures of this kind is in keeping with the cross-linguistic tendency for “combinations of words and morphemes that occur together very frequently […] to be stored and processed in one chunk” (Bybee 2003: 617), thus becoming particularly resistant to change. Residual usage of do-less negatives disappeared rapidly after 1800, though occasional instances of such semi-idiomatic expressions occur even in the last decades of the 19th century and in the early 20th century, as in (18) and (19). (18)
Whether he uses tobacco thus openly as a friendly fumigative only I know not (1897 Daily News 13 Feb. 6/4; OED, s.v. fumigative a. and n.)
(19)
It must have been some old sacred language – Phoenician, Sabæan, I know not what – which had survived in the rite (1910 Buchan, Prester John xi. 183; OED, s.v. Sabæan, Sabean a. and n.)
3.4 Shifts in the complementation system The system of complementation has been subject to important changes from Old English times to the present day, some of them consolidating in our period and culminating in an important rearrangement of the system, which has recently been labelled the “Great Complement Shift” (Rohdenburg 2006: 143). (See also Seoane, Chapter 39, Section 5.) As far as finite complementation is concerned, our period sees the continuation of the “long-term trend in English” (Denison 1998: 256) to replace finite complements by non-finite clauses, mostly by infinitives. However, recent research has shown that, since that-clauses are easier to process than non-finite structures, they tend to be retained in complex cognitive contexts, such as insertions of intervening material or negated complements (cf. Rohdenburg 1995, 2006). Complexity factors have also been attributed an important role in the variation between infinitives and -ing complements, as will be commented on below. Apart from the decline of that-clauses, the grammar of finite complementation has not changed much over the last three centuries. In our period the major complementizers continue to be that and zero, but the advance of the latter, which had started in Middle English and continued at a rapid pace in Early Modern English, experienced a halt in the “norm-loving” 18th century (Rissanen 1991: 288; cf. also Denison 1998: 259), especially in formal writings (cf. Finegan and Biber 1995). Another outstanding aspect in the domain of finite complementation is the partial revival of the so-called mandative subjunctive in complements to verbs of commanding, requesting
54 Late Modern English: Syntax and the like (i.e. manipulative predicates). The use of the subjunctive had been decreasing steadily ever since the Middle English period in all types of subordinates, complement clauses included. From the 19th century onwards, however, there is a slight recovery of the mandative subjunctive. While in American English the subjunctive has always been the default option in complements to manipulative predicates, in British English its re-emergence seems to be closely linked with the written language and particularly with legal texts (Denison 1998: 262–264; Huddleston and Pullum et al. 2002: 995). It would be worth investigating whether prescriptivism, which has been claimed to have had an effect on the resurgence of the subjunctive in adverbial clauses in the late 18th and early 19th centuries (Auer 2006), could also have played a role in its revival in complementation structures. While we can speak of relative stability in the area of finite complementation, the realm of non-finite complementation experienced “fundamental and rapid changes” in our period (Mair 2003: 329), some of them still underway in the present day. It should therefore come as no surprise that non-finite complementation has become an active research topic over the last few years. Perhaps the most relevant change in this area is the replacement of to-infinitives by -ing complements, which started in Late Middle English, but gained momentum from the 19th century onwards (Strang 1970: 100). Gerunds have their origin in Old English action nouns in -ing/-ung, and behaved as nouns until Late Middle English, when they started to acquire verbal properties (among them the ability to take objects, to take a subject in the oblique case, to be modified by adverbs restricted to verbs, and to be negated by means of not). In the Late Modern English period -ing forms range from the clearly nominal (example 20) to the clearly verbal (example 21), featuring also hybrid structures, which had become frequent from the late 17th century onwards. These hybrids were particularly common in positions where verbal gerunds were not licensed at first, notably as preverbal subjects (example 22). There is no agreement as to the function of the definite article in such structures: Fanego (2007b: 192) argues that it is used to identify the -ing clause as a complement (i.e. it serves a complementizer function), while, according to De Smet (2008: 64–67), it retains its article-like functionality. Hybrids have declined in frequency since the late 18th century and, although they had virtually disappeared by the late 19th century, occasional examples can still be found in 20th-century English (Aarts 2007: 229). (20)
In order to this I secretly employ’d my Confessor, a very good Ecclesiastick, to propose the purchasing of my Estate and Houses, or rather Palaces […] to my Wife’s Relations; (1739 Aubin, Count Albertus Chap. 7, 278; ECF)
(21)
[…] he has, or I fancy he has, all the insolence of a happy rival; ’tis unjust, but I cannot avoid hating him; (1769 Brooke, Emily Montague Vol. 1, Letter 27, 149; ECF)
(22)
[…] he chose to make the first Declaration to herself; the gaining her Affections being the material Point, he considered all others of little Consequence. (1725 Haywood, Fatal Secret 271; Fanego 2007b: 192)
879
880
VI Late Modern English It has been surmised that true hybrids, which “are characterized by an equal number of properties from two categories”, are difficult to process and are therefore generally avoided in languages (Aarts 2007: 229, 233). In addition to this processing factor, the demise of hybrid structures may have been influenced by other determinants, among them the increasing polarization of the gerund into a purely verbal type (reinforced by the dramatic increase of the progressive, discussed in Section 3.1 above) and a purely nominal type (supported by the steady growth of action nouns like destruction, blockage, and betrayal, cf. van der Wurff 1993), and the condemnation of prescriptive grammarians (cf. Visser 1963–73: Sections 1124, 1040; Fanego 2006). Prescriptivism may also be held responsible for the promotion of genitive over objective noun phrases as subjects of gerunds (e.g. I hate his/him telling sexist jokes) in educated speech until the 20th century (Denison 1998: 269; Mair 2006a: 223; Mugglestone 2006: 285). From the beginning of their use, verbal gerunds have encroached upon the domain of to-infinitival complements. There is general agreement that the type of matrix verb plays an essential role in this change. The model seems to be one of lexical diffusion, with verbs of avoidance (avoid, forbear, etc.) being the first ones to adopt the newer construction, which then spread to other verbs of negative implication like verbs of refusal (decline, deny, refuse, etc.) (cf. Fanego 2007b: 170; also Rudanko 2000: Chapter 7). The variation between to-infinitives and gerunds after verbs of negative import is influenced by stylistic tendencies like horror aequi (Rohdenburg 2003: 236), as gerunds tend to be avoided if the matrix predicate is itself an -ing form (Rudanko 2000: 111–112). In Late Modern English -ing complements advanced to verbs like remember when used retrospectively. The typical construction until the second half of the 18th century was the perfect infinitive (De Smet and Cuyckens 2007: 192), as in example (23). From that time onwards, perfect infinitives had to compete with -ing complements (example 24), which were felt to be particularly suitable for that function, perhaps because they were “indifferent to time distinctions” due to their original nominal nature (Fanego 2007b: 175). (23)
[…] and she pointed to the high gallery where I had seen her going out on that same first day, and told me she remembered to have been up there, and to have seen me standing scared below. (1861 Dickens, Great Expectations Vol. II, Ch. X, 166; NCF)
(24)
I perfectly remember carrying back the Manuscript you mention and delivering it to Lord Oxford. (1740 Lady Mary Wortley Montagu, Complete Letters [Halsband], 354; Fanego 2007b: 175)
Although the gerund with retrospective verbs had to face the rejection of grammarians in the early 19th century (Visser 1963–73: Section 1777), it had ousted the perfect infinitive almost completely by 1900 (but cf. Mair 2006a: 226 and De Smet and Cuyckens 2007: 193, who report the existence of sporadic examples of the older construction as late as the second half of the 20th century). The decline of the perfect infinitive and the concomitant increase in the use of -ing complements with retrospective verbs seems to have been delayed in certain environments, such as (i) cognitively complex contexts, like extractions and insertions of intervening material (cf. the so-called Complexity Principle, Rohdenburg 2003); (ii) in cases of horror aequi, where two -ing forms
54 Late Modern English: Syntax would be adjacent (Vosberg 2003); and (iii) when the complement was a verb of perception or encounter (De Smet and Cuyckens 2007: 193). Once well established with remember, -ing complements could spread to other semantically related verbs like forget (19th century) and recall (20th century) (Fanego 2007b: 175). Another group of verbs which takes the gerund much more commonly in the Late Modern English period is that of the aspectuals (begin, cease, commence, etc.), while conatives (try) and emotive verbs (enjoy, hate, like) are first found with the gerund in our period (Fanego 2007b: 178). In addition to the type of matrix predicate, research by Fanego (2004) has shown that the function of the complement has a bearing on the spread of gerund clauses. Originally verbal gerunds appeared almost exclusively after prepositions (cf. also De Smet 2008), an environment in which the occurrence of the to-infinitive was virtually precluded (cf. Fanego 2007b: 170n. 9), then moving to object function, and finally spreading to subject function. Unlike other complement-types, gerunds in subject function have always been resistant to extraposition (cf. Kaltenbo¨ck 2004 on Present-day English; Fanego 2010 for a historical overview). This behavior might be related to their original nominal nature, as nouns cannot be extraposed. The gradual spread of gerunds to the preverbal subject function, which takes place from the second half of the 18th century (cf. Fanego 2010), seems to have been facilitated by the existence of hybrids introduced by the definite article, as in example (22) above. According to Fanego (2010), the selection of the gerund over the infinitive in preverbal position seems to be determined by the length of the complement, since preverbal gerunds are typically lighter than infinitives. Finally, another non-finite complement-type undergoing changes throughout the Late Modern English period is the for … to-construction. This pattern originates in Middle English structures of the type it is good/bad/shameful [for NP][to X] (Fischer 1992: 330– 331), where the for-phrase has a benefactive reading and does not form a constituent with the following to-infinitive. The genuine for…to-construction first appeared in extraposed subject function (example 25), where it was well established by the beginning of the Early Modern English period. Unlike in benefactive structures, in the for … toconstruction the for-phrase and the infinitive operate as a single unit, with the for-phrase analysed as the subject of the to-infinitive. Expansion to the object function, as in (26), is relatively late (Cuyckens and De Smet 2007: 99; De Smet 2007: 79), and does not seem to have really taken off until the 20th century (Mair 2006b: 125). (25)
By this tale men may se it is no wysedome for a man to attempte a meke wom~n pacye~ce to far. (1526 A Hundred Mery Talys 115, Helsinki Corpus; Cuyckens and De Smet 2007: 94)
(26)
As soon as ever it’s possible, we’ll arrange for you to live with someone who will preserve appearances. (1893 Gissing, The Odd Woman; De Smet 2007: 85)
3.5 Relativizers Another interesting area of regulatory change in our period is that of relativization. Once the wh-forms (which, whom, whose, and who) came to be used as relative markers during Middle English, in addition to that and zero, the inventory of English relativizers
881
882
VI Late Modern English had already become identical to that of the present day. The modern period, in particular the 18th century, saw the imposition of a number of constraints on the patterns of distribution of these forms, such restrictions being still operative in the contemporary standard language. One of these regulating tendencies concerned the distinction between human who and non-human which along the so-called animacy parameter. When which emerged in Middle English as a relativizer, it could be used with reference to both personal and non-personal antecedents, and this usage continued in the Early Modern English period (cf. Johansson, Chapter 49). Eighteenth-century grammarians vehemently condemned the “misuse” of human which (cf. example 27 below), which progressively became less common, though it could occur in non-standard English even by the end of the century (cf. Austin 1985: 18–19). (27)
I was at Mr. Barrons when Mr. Paynter wich is my Master Came ther (Elizabeth Clift’s letters; Austin 1985: 26)
The use of the invariable relativizer that was subjected to regulation along the information parameter also during Late Modern English. In his famous “Humble Petition of Who and Which” (1711, number 78 of the Spectator), Addison complains about the excessive use of that characteristic of the 17th century. The functional overload of that, which served also as a complementizer and a demonstrative, may well have supported the tendency to confine it to restrictive (or defining) clauses and to promote the use of the “maximally distinctive and minimally ambiguous” wh-relativizers (Beal 2004: 76) since the 18th century (cf. also Rissanen 1999: 295; Beal 2004: 113). Ball (1996: 248–251) convincingly shows that who and which rapidly gained ground at the expense of invariant that in the period 1700–1800, and that the expansion of the wh-forms continued somewhat more gradually in the 19th century. Statistical and regulatory changes also affected the use of the zero relativizer at this time. Visser (1963–73: Section 630) reports how “in the course of the eighteenth and nineteenth centuries a remarkable decline in the currency of the zero-construction becomes perceptible”, zero becoming progressively associated with informal style. As today, the omission of the relativizer in Late Modern English seems to have been particularly disfavored in the subject function, as in (28). (28)
O there is that disagreeable Lover of mine Sir Benjamin Backbite Ø has just call’d at my Guardian’s (1777 Sheridan, School for Scandal I.i.363.22; Denison 1998: 281)
4 Summary We hope that the foregoing discussion has shown that the syntax of Late Modern English deserves detailed attention, because it is in this period that some fundamental long-term changes culminate and that the foundations of important Present-day English usages are to be found. In the last few years various monographs (overviews of 18th and 19th century English like Bailey 1996; Go¨rlach 1999, 2001; Beal 2004; Tieken-Boon van Ostade 2009, among others), as well as studies of particular grammatical features (e.g. Smitterberg 2005; Fanego 2007b), collections of papers (e.g. Dossena and Jones 2003; Kyto¨ et al. 2006; Pe´rez-Guerra et al. 2007; Tieken-Boon van Ostade and van der Wurff
54 Late Modern English: Syntax 2009; Hickey 2010), and conferences (e.g. The Late Modern English Conferences 1–4) have succeeded in partially making up for the traditional neglect of the period. We believe, however, that there is still room for further research, both in the areas discussed in the present chapter and regarding the other issues mentioned in the introduction. The wealth of materials covering the period (especially corpora), which have been made available over the last decade, will undoubtedly facilitate the task of those willing to bridge the descriptive gap between the Early Modern English and Present-day English periods. Acknowledgments: The authors would like to dedicate this chapter to the late Richard Hogg. Bas Aarts would like to thank Mariangela Spinillo and Sean Wallis for permission to use material based on joint work presented at the Directions in English Language Studies conference in Manchester in 2006. For generous financial support, Marı´a Jose´ Lo´pezCouso and Bele´n Me´ndez-Naya are grateful to the European Regional Development Fund and the following institutions: Spanish Ministry for Science and Innovation (Grants HUM2007-60706 and FF12011-26693-C02-01) and Autonomous Government of Galicia (Directorate General for Scientific and Technological Promotion, grant CN 2011/011).
5 References Aarts, Bas. 2007. Syntactic Gradience: The Nature of Grammatical Indeterminacy. Oxford: Oxford University Press. Aarts, Bas, Joanne Close, and Sean Wallis. 2010. Recent changes in the use of the progressive construction in English. In: Bert Cappelle and Naoaki Wada (eds.), Distinctions in English Grammar, offered to Renaat Declerck, 148–167. Tokyo: Kaitakusha. Aarts, Bas and Sean Wallis. 2006. The Diachronic Corpus of Present-Day Spoken English (DCPSE). Survey of English Usage, University College London. http://www.ucl.ac.uk/ english-usage/projects/dcpse/ Arnaud, Rene. 1983. On the progress of the progressive in the private correspondence of famous British people (1800-1880). In: Sven Jacobson (ed.), Papers from the Second Scandinavian Symposium on Syntactic Variation: Stockholm, May 15–16, 1982, 83–94. Stockholm: Almqvist and Wiksell. (Also published in Language Variation and Change 10: 123–152.) Auer, Anita. 2006. Precept and practice: The influence of prescriptivism on the English subjunctive. In: Dalton-Puffer et al. (eds.), 33–53. Austin, Frances O. 1985. Relative which in late 18th century usage: The Clift family correspondence. In: Roger Eaton, Olga Fischer, Willem Koopman, and Frederike van der Leek (eds.), Papers from the 4th International Conference on English Historical Linguistics, 15–29. Amsterdam/Philadelphia: John Benjamins. Bailey, Richard W. 1996. Nineteenth-century English. Ann Arbor: The University of Michigan Press. Ball, Catherine N. 1996. A diachronic study of relative markers in spoken and written English. Language Variation and Change 8: 227–258. Beal, Joan C. 2004. English in Modern Times. London: Arnold. Biber, Douglas and Edward Finegan 1990–93/2002/2007/2010. A Representative Corpus of Historical English Registers (ARCHER). Version 3.1. Consortium of fourteen universities, see http:// www.llc.manchester.ac.uk/research/projects/archer/. Bybee, Joan. 2003. Mechanisms of change in grammaticization: The role of frequency. In: Brian D. Joseph and Richard D. Janda (eds.), The Handbook of Historical Linguistics, 602–623. Oxford: Blackwell. Chadwyck-Healey. 1996–2011. Literature Online (LION). http://lion.chadwyck.com/ Cuyckens, Hubert and Hendrik De Smet. 2007. For … to-infinitives from Early to Late Modern English. In: Pe´rez-Guerra et al. (eds.), 77–103.
883
884
VI Late Modern English Dalton-Puffer, Christiane, Dieter Kastovsky, Nikolaus Ritt, and Herbert Schendl (eds.). 2006. Syntax, Style and Grammatical Norms: English from 1500–2000. Bern: Peter Lang. De Smet, Hendrik. 2007. For … to-infinitives as verbal complements in Late Modern and PresentDay English: Between motivation and change. English Studies 88(1): 67–94. De Smet, Hendrik. 2008. Functional motivations in the development of nominal and verbal gerunds in Middle and Early Modern English. English Language and Linguistics 12(1): 55–102. De Smet, Hendrik and Hubert Cuyckens. 2007. Diachronic aspects of complementation: Constructions, entrenchment and the matching problem. In: Christopher M. Cain and Geoffrey Russom (eds.), Studies in the History of the English Language III, 187–214. Berlin/New York: Mouton de Gruyter. Denison, David. 1993. English Historical Syntax. Harlow: Longman. Denison, David. 1998. Syntax. In: Suzanne Romaine (ed.), The Cambridge History of the English Language. Vol. IV. 1776-1997, 92–329. Cambridge: Cambridge University Press. Denison, David, Graeme Trousdale, and Linda van Bergen. 1994. A Corpus of late Modern English Prose. http://www.llc.manchester.ac.uk/temp/lel/david-denison/lmode-prose/. Dennis, Leah. 1940. The progressive tense: Frequency of its use in English. Publications of the Modern Language Association of America 55: 855–865. Dossena, Marina and Charles Jones (eds.). 2003. Insights into Late Modern English. Bern: Peter Lang. Fanego, Teresa. 2004. On reanalysis and actualization in syntactic change: The rise and development of English verbal gerunds. Diachronica 21: 5–55. Fanego, Teresa. 2006. The role of language standardization in the loss of hybrid gerunds in Modern English. In: Leiv Egil Breivik, Sandra Halverson, and Kari E. Haugland (eds.), These things write I vnto thee …: Essays in Honour of Bjorg Baekken, 93–110. Oslo: Novus. Fanego, Teresa. 2007a. COLMOBAENG: A Corpus of Late Modern British and American English Prose 1700–1879. University of Santiago de Compostela. Fanego, Teresa. 2007b. Drift and development of sentential complements in British and American English from 1700 to the present day. In: Pe´rez-Guerra et al. (eds.), 161–235. Fanego, Teresa. 2010. Variation in sentential complements in eighteenth- and nineteenth-century English: A processing-based explanation. In: Hickey (ed.), 200–220. Finegan, Edward and Douglas Biber. 1995. That and zero complementisers in Late Modern English: Exploring ARCHER from 1650–1990. In: Bas Aarts and Charles Meyer (eds.), The Verb in Contemporary English, 241–257. Cambridge: Cambridge University Press. Fischer, Olga. 1992. Syntax. In: Norman Blake (ed.), The Cambridge History of the English Language. Vol. II. 1066-1476, 207–408. Cambridge: Cambridge University Press. Fleisher, Nicholas. 2006. The origin of passive get. English Language and Linguistics 10(2): 225–252. Francis, W. Nelson and Henry Kucˇera. 1979 [1964]. A Standard Corpus of Present-day Edited English (Brown Corpus). Brown University. In: ICAME Collection of English Language Corpora. (CD-ROM). 2nd edn., 1999. Knut Hofland, Anne Lindebjerg, and Jørn Thunestvedt (eds.), The HIT Centre, University of Bergen, Norway. For manual, see http://khnt.aksis.uib. no/icame/manuals/brown/INDEX.HTM Givo´n, Talmy and Lynne Yang. 1994. The rise of the English GET-passive. In: Barbara Fox and Paul J. Hopper (eds.), Voice, Form and Function, 119–149. Amsterdam/Philadelphia: John Benjamins. Go¨rlach, Manfred. 1999. English in Nineteenth-Century England. Cambridge: Cambridge University Press. Go¨rlach, Manfred. 2001. Eighteenth-Century English. Heidelberg: Winter. Greenbaum, Sidney and Gerald Nelson. 1998. The British Component of the International Corpus of English (ICE-GB). Survey of English Usage, University College London. http://www.ucl.ac. uk/english-usage/projects/ice-gb/index.htm Greenbaum, Sidney and Jan Svartvik. 1990. The London-Lund Corpus of Spoken English (LLC). In: ICAME Collection of English Language Corpora. (CD-ROM). 2nd edn., Knut Hofland, Anne Lindebjerg, and Jørn Thunestvedt (eds.), The HIT Centre, University of Bergen, Norway. For manual, see http://khnt.aksis.uib.no/icame/manuals/LONDLUND/INDEX.HTM
54 Late Modern English: Syntax Gronemeyer, Claire. 1999. On deriving complex polysemy: The grammaticalization of get. English Language and Linguistics 3(1): 1–39. Hawley, Judith, Tom Keymer, and John Mullan. 1996. Eighteenth-century Fiction. ChadwyckHealey Ltd. http://collections.chadwyck.com/home/home_c18f.jsp Hickey, Raymond (ed.). 2010. Eighteenth-century English. Ideology and Change. Cambridge: Cambridge University Press. Huddleston, Rodney and Geoffrey Pullum et al. 2002. The Cambridge Grammar of the English Language. Cambridge: Cambridge University Press. Hundt, Marianne. 2001. What corpora tell us about the grammaticalisation of voice in getconstructions. Studies in Language 25(1): 49–88. Hundt, Marianne. 2004. Animacy, agentivity, and the spread of the progressive in Modern English. English Language and Linguistics 8(1): 47–69. Johnson, Samuel. 1979 [1755]. A Dictionary of the English Language. First published London: W. Strahern. Reprinted London: Times Books. Kaltenbo¨ck, Gunther. 2004. Using non-extraposition in spoken and written texts. A functional perspective. In: Karin Aijmer and Anna-Brita Stenstro¨m (eds.), Discourse Patterns in Spoken and Written Corpora, 219–242. Amsterdam/Philadelphia: John Benjamins. Karlin, Daniel and Tom Keymer. 2000. Nineteenth-century Fiction. Chadwyck-Healey Ltd. http:// collections.chadwyck.com/marketing/home_c19f.jsp Kranich, Svenja. 2009. Interpretative progressives in Late Modern English. In: Tieken-Boon van Ostade and van der Wurff (eds.), 331–357. Krug, Manfred. 2000. Emerging English Modals. A Corpus-based Study of Grammaticalization. Berlin/New York: Mouton de Gruyter. Kyto¨, Merja. 1997. Be/have + past participle: The choice of the auxiliary with intransitives from Late Middle to Modern English. In: Matti Rissanen, Merja Kyto¨, and Kirsi Heikkonen (eds.), English in Transition. Corpus-based Studies in Linguistic Variation and Genre Styles, 17–85. Berlin/New York: Mouton de Gruyter. Kyto¨, Merja and Juhani Rudanko forthc. Corpus of Nineteenth-century English (CONCE). Uppsala University and University of Tampere. Kyto¨, Merja, Mats Ryde´n, and Erik Smitterberg (eds.). 2006. Nineteenth-Century English: Stability and Change. Cambridge: Cambridge University Press. Leech, Geoffrey N. and Stig Johansson. 1978. The Lancaster-Oslo/Bergen Corpus of British English (LOB) In: ICAME Collection of English Language Corpora. (CD-ROM). 2nd edn., 1999. Knut Hofland, Anne Lindebjerg, and Jørn Thunestvedt (eds.), The HIT Centre, University of Bergen, Norway. For manual, see http://khnt.aksis.uib.no/icame/manuals/lob/INDEX. HTM Lowth, Robert. 1979. [1775] A Short Introduction to English Grammar. Scholars’ facsimiles and reprints. New York: Delmar. Mair, Christian. 1998. The Freiburg-LOB Corpus of British English (F-LOB). In: ICAME Collection of English Language Corpora. (CD-ROM). 2nd edn., 1999. Knut Hofland, Anne Lindebjerg, and Jørn Thunestvedt (eds.), The HIT Centre, University of Bergen, Norway. For manual, see http://khnt.aksis.uib.no/icame/manuals/flob/INDEX.HTM Mair, Christian. 1999. The Freiburg-Brown Corpus of American English (Frown). In: ICAME Collection of English Language Corpora. (CD-ROM). 2nd edn., 1999. Knut Hofland, Anne Lindebjerg, and Jørn Thunestvedt (eds.), The HIT Centre, University of Bergen, Norway. For manual, see http://khnt.aksis.uib.no/icame/manuals/frown/INDEX.HTM Mair, Christian. 2003. Gerundial complements after begin and start: Grammatical and sociolinguistic factors, and how they work against each other. In: Rohdenburg and Mondorf (eds.), 329–345. Mair, Christian. 2006a. Nonfinite complement clauses in the nineteenth century: The case of remember. In: Merja Kyto¨ et al. (eds.), 215–228. Mair, Christian. 2006b. Twentieth-century English: History, Variation and Standardization. Cambridge: Cambridge University Press.
885
886
VI Late Modern English Mair, Christian and Marianne Hundt. 1995. Why is the progressive becoming more frequent in English? Zeitschrift fu¨r Anglistik und Amerikanistik 43: 111–122. Mair, Christian and Geoffrey Leech. 2006. Current changes in English syntax. In: Bas Aarts and April McMahon (eds.), The Handbook of English Linguistics, 318–342. Malden MA: Blackwell. Milic, Louis T. 1995. The Century of Prose Corpus: A half-million word historical data base. Computers and the Humanities 29: 327–337. Mugglestone, Lynda. 2006. English in the nineteenth century. In: Mugglestone (ed.), 274–304. Mugglestone, Lynda (ed.). 2006. The Oxford History of English. Oxford: Oxford University Press. Nesselhauf, Nadja. 2007. The spread of the progressive and its ‘future’ use. English Language and Linguistics 11(1): 191–207. Nu´n˜ez-Pertejo, Paloma. 2004. The Progressive in the History of English with Special Reference to the Early Modern English Period: A Corpus-based Study. Mu¨nchen: Lincom Europa. Nu´n˜ez-Pertejo, Paloma. 2007. Aspects of the use of the progressive in the eighteenth century. In: Pe´rez-Guerra et al. (eds.), 359–382. Pe´rez-Guerra, Javier, Dolores Gonza´lez-Alvarez, Jorge L. Bueno-Alonso, and Esperanza RamaMartı´nez (eds.). 2007. “On varying language and opposing creed”. New Insights into Late Modern English. Bern: Peter Lang. Rissanen, Matti. 1991. On the history of that/zero as object clause links in English. In: Karin Aijmer and Bengt Altenberg (eds.), English Corpus Linguistics: Studies in Honour of Jan Svartvik, 272–289. London: Longman. Rissanen, Matti. 1999. Syntax. In: Roger Lass (ed.), The Cambridge History of the English Language. Vol. III. 1476-1776, 187–331. Cambridge: Cambridge University Press. Rissanen, Matti, Merja Kyto¨, Leena Kahlas-Tarkka, Matti Kilpio¨, Saara Nevanlinna, Irma Taavitsainen, Terttu Nevalainen, and Helena Raumolin-Brunberg. 1991. The Helsinki Corpus of English Texts. In: ICAME Collection of English Language Corpora (CD-ROM), 2nd edn., 1999. Knut Hofland, Anne Lindebjerg, and Jørn Thunestvedt (eds.), The HIT Centre, University of Bergen, Norway. For manual, see http://khnt.hit.uib.no/icame/manuals/hc/index.htm Rohdenburg, Gu¨nter. 1995. On the replacement of finite complement clauses by infinitives in English. English Studies 76: 367–388. Rohdenburg, Gu¨nter. 2003. Cognitive complexity and horror aequi as factors determining the use of interrogative clause linkers in English. In: Rohdenburg and Mondorf (eds.), 205–249. Rohdenburg, Gu¨nter. 2006. The role of functional constraints in the evolution of the English complement system. In: Dalton-Puffer et al. (eds.), 143–166. Rohdenburg, Gu¨nter and Britta Mondorf (eds.). 2003. Determinants of Grammatical Variation in English. Berlin/New York: Mouton de Gruyter. Rudanko, Juhani. 2000. Corpora and Complementation. Tracing Sentential Complementation Patterns of Nouns, Adjectives and Verbs over the Last Three Centuries. Lanham, MD: University Press of America. Ryde´n, Mats and Sverker Brorstro¨m. 1987. The be/have Variation with Intransitives in English, with Special Reference to the Late Modern Period. Stockholm: Almqvist and Wiksell. Simpson, John (ed.). 2000–. The Oxford English Dictionary. 3rd edn. online. Oxford University Press. www.oed.com/. Smith, Nicholas. 2005. A Corpus-based Investigation of Recent Change in the Use of the Progressive in British English. Ph.D. dissertation, Department of Linguistics and English Language, Lancaster University. Smitterberg, Erik. 2005. The Progressive in 19th-century English: A Process of Integration. Amsterdam: Rodopi. Smitterberg, Erik. 2008. The progressive and phrasal verbs: Evidence of colloquialisation in nineteenth-century English. In: Terttu Nevalainen, Irma Taavitsainen, Pa¨ivi Pahta, and Minna Korhonen (eds.), The Dynamics of Linguistic Variation: Corpus Evidence on English Past and Present, 269–289. Amsterdam/Philadelphia: John Benjamins. Strang, Barbara M. H. 1970. A History of English. London: Methuen.
55 Late Modern English: Semantics and lexicon
887
Strang, Barbara M. H. 1982. Some aspects of the history of the be + -ing construction. In: John Anderson (ed.), Language Form and Linguistic Variation: Papers Dedicated to Angus McIntosh, 427–474. Amsterdam/Philadelphia: John Benjamins. Tieken-Boon van Ostade, Ingrid. 1987. The Auxiliary do in Eighteenth-Century English: A Sociohistorical Linguistic Approach. Dordrecht: Foris. Tieken-Boon van Ostade, Ingrid. 2006. English at the onset of the normative tradition. In: Lynda Mugglestone (ed.), 240–273. Tieken-Boon van Ostade, Ingrid. 2009. An Introduction to Late Modern English. Edinburgh: Edinburgh University Press. Tieken-Boon van Ostade, Ingrid and Wim van der Wurff (eds.). 2009. Current Issues in Late Modern English. Bern: Peter Lang. Traugott, Elizabeth Closs. 1972. The History of English Syntax. A Transformational Approach to the History of English Sentence Structure. New York: Holt, Rinehart and Winston. Traugott, Elizabeth Closs. 1992. Syntax. In: Richard M. Hogg (ed.), The Cambridge History of the English Language. Vol. I. The beginnings to 1066, 168–289. Cambridge: Cambridge University Press. van der Wurff, Wim. 1993. Gerunds and their objects in the Modern English period. In: Jaap van Marle (ed.), Historical Linguistics 1991, 363–375. Amsterdam/Philadelphia: John Benjamins. Visser, Frederikus Theodorus. 1963–73. An Historical Syntax of the English Language. Leiden: E. J. Brill. Vosberg, Uwe. 2003. Cognitive complexity and the establishment of -ing constructions with retrospective verbs in Modern English. In: Dossena and Jones (eds.), 197–220. Wright, Susan. 1994. The mystery of the modal progressive. In: Dieter Kastovsky (ed.), Studies in Early Modern English, 467–485. Berlin/New York: Mouton de Gruyter. Wright, Susan. 1995. Subjectivity and experiential syntax. In: Dieter Stein and Susan Wright (eds.), Subjectivity and Subjectivisation: Linguistic Perspectives, 151–172. Cambridge: Cambridge University Press.
Bas Aarts, London (UK), Marı´a Jose´ Lo´pez-Couso, Santiago de Compostela (Spain), and Bele´n Me´ndez-Naya, Santiago de Compostela (Spain)
55 Late Modern English: Semantics and lexicon 1 2 3 4 5 6 7
Introduction A new era: new vocabulary and new meanings Obsolescence and proscription Tracing the roots of English Landmarks in lexicography Summary References
Abstract This chapter spans little more than two centuries, going from the early 18th century (with the rise of the age of “politeness”), through the Victorian age, to the beginning of the 20th century. These are crucial decades, in which lexis and semantics were deeply affected by changes in society and varying epistemological approaches to the past that had an impact Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 887–900
888
VI Late Modern English on the study of etymology, onomastics, and lexicography. In the opening section attention is given to new formations, calques, and loanwords that appear to have been recorded in LModE for the first time; the section will also focus on semantic change. There follows a discussion of those lexical items that either became obsolete in LModE, or the use of which was explicitly discouraged in grammars, dictionaries, and textbooks of the time. Finally, contemporary scholarly attitudes to English vocabulary are discussed, including an overview of the main dictionaries that were published in the 18th and 19th centuries.
1 Introduction The time scope of this chapter goes from the early 18th century (with the rise of the age of “politeness”), through the Victorian age, to the beginning of the 20th century. These are crucial decades in which lexis and semantics were deeply affected by changes in society: e.g. the Industrial Revolution and the effects it had on legal, social, and economic spheres. Such changes were also brought about by political and historical factors like the French Revolution, the Napoleonic Wars, the rise of the British Empire, and the First World War; also varying epistemological approaches to the past had an impact on the study of etymology, onomastics, and lexicography. As a matter of fact, the history of vocabulary places itself at the intersection of the internal and the external histories of a language (Kastovsky 2006: 270). In vocabulary, variation and change depend on both internal factors, such as the increasing or decreasing productivity of word formation processes, and external ones, such as the need to express new concepts dictated by changes in external reality. In addition, social conditioning may encourage or discourage strategies like calques and borrowings, or it may favor or disfavor lexical choices more closely associated with one language family than another (for instance, it is the case of “Anglo-Saxon” monosyllabic words being (dis)preferred over “Latinate”, polysyllabic ones on account of the supposedly greater colloquial traits of the former). Finally, vocabulary is the branch in which the creativity of speakers can express itself most freely (Algeo 1998: 58–59): users can and do manipulate lexis for jocular purposes, producing formations that may or may not be adopted by other speakers and may or may not survive the test of time even in the repertoire of the same speaker; or they may create new items meant to surprise recipients, such as we see in the case of advertisements deliberately shaping new slogans and names of products out of well-known lexical items. Indeed this was one of the processes at work in the 19th century, when new forms of commercial communication began to spread (Bailey 1996: 21). Bailey (1996: 321) also stresses the importance of “words” and their almost symbolic status as emblems of erudition and power: for instance, the rise of parlor games centered on words shows the fascination they held for speakers, especially in the middle classes. Vocabulary thus acquired importance as the first vehicle by means of which change – indeed, “progress” (Bailey 1996: 2) – manifested itself. According to the Oxford English Dictionary (OED), over 101,000 entries were first cited in the years spanning 1700 and 1918; in particular, findings may be broken down as in Table 55.1.
55 Late Modern English: Semantics and lexicon
889
Table 55.1: Items by first citation date in the OED Time span
New entries
1700–1750 1751–1800 1801–1850 1851–1900 1901–1918 Total
10920 12665 29786 40172 7852 101395
The rise is perhaps not so dramatic if we consider that in the two centuries spanning 1500–1700 c.85,000 new entries were recorded; in addition, it should be borne in mind that these findings may be a function of what documents were scanned for entries. If we wish to study lexical and semantic change, then, relying on dictionaries may not be sufficient, as these were naturally compiled with a rather different epistemological agenda from a contemporary one: early dictionaries, for instance, mainly aimed to provide glossaries and lists of “hard words”, instead of recording new usage. Nor can dictionaries be expected to provide accurate information on orthographic variation, as they recorded the norm that was codified, but not necessarily what was in current usage; indeed, Mugglestone (2006: 279) stresses the importance of studying authentic manuscripts, as printers felt entitled to “standardize” spelling, and the language we access today in edited texts may not reflect actual usage – see also Lass (2004). Bearing in mind this caveat, it would perhaps be interesting then to base investigations on the numerous corpora currently available or in preparation – see Dossena (2004), Fitzmaurice (2004, 2006), Tieken Boon van Ostade (2005), Dury (2006), Kyto¨ et al. (2006), van Bergen and Denison (2007), Dollinger (2008), Davies (2010–), and Corbett et al. (2010). This would allow us to see to what extent the information provided in dictionaries does match real usage beyond the texts taken into consideration by lexicographers and their readers. However, it is nonetheless significant that nearly 40% of these new entries were first recorded in the second half of the 19th century, when social, technological, and political innovation led to lexical change. The kind of lexical and semantic variation that can be identified in these decades and the ways in which scholars and lexicographers approached them will be the object of the next sections.
2 A new era: new vocabulary and new meanings It would be beyond the aims of this chapter to rehearse the extensive findings already offered by Algeo (1998), Go¨rlach (1999, 2001), Beal (2004), Mugglestone (2000, 2006) and Kastovsky (2006) in relation to processes of word formation, calques and borrowings in the period under discussion. What appears to be more interesting is the kind of trends that these new items appear to outline in relation to the areas in which lexical and semantic innovation took place. We have already mentioned some of the fields in which new lexis arose: science, technology, but also advertising, among others. To these, we can add the legal novelties introduced by the copyright and trademark laws, Acts concerning postage, compulsory education and voting; the spread of English as a lingua franca in the many territories of the British Empire also played a role both in the need for new vocabulary capable of
890
VI Late Modern English translating and representing the reality of those territories and in the need for new dictionaries that could be used with a whole new category of learners. Within these branches, it then appears particularly intriguing to see how indebted the present-day world is to discoveries, and therefore vocabulary, first recorded in the 19th century. This is especially true in disciplines like geology: an item that today has entered popular use, Jurassic, dates from 1831, while Triassic from 1841, the same year in which dinosaur appears to have been first recorded. According to the OED, palaeontology is first recorded in 1836, while bacteria are first discussed in 1847. As for silicon, without which so many appliances of today would be unthinkable, it was first named in 1817, replacing silicium. Especially in science, new lexical items were derived from the names of the people who had theorized the phenomena to which the new vocabulary applied: it is the case of volt (1873, from the name of the Italian physicist and scientist, Alessandro Volta, who made very early experiments with electricity in the late 18th century), ampe`re (1881, from the name of a French electrician), watt (1882, thus called in honor of James Watt), ohm (1861, from the name of the German physicist who determined mathematically the law of the flow of electricity) and faraday (used attributively since 1886, from the surname of Michael Faraday). A similar process had already been at work in the late 18th century when, for instance, dolomite (1794) was coined from Dolomieu, the surname of the French geologist and mineralogist who had discovered the characteristics of this stone. Nor was it just nouns that were affected: the verb to galvanize, first used metaphorically in 1853 by Charlotte Bronte¨, but first recorded in 1802, derives from galvanism, ‘electricity developed by chemical action’, which in turn is based on the name of Luigi Galvani, another Italian scientist who first described such phenomena in 1792. In medicine, the influence of French was to continue in such new formations as meningitis (1824), modelled on French me´ningite´ (1793), and chloroform (1834), first described by Simpson as a powerful anaesthetic in 1847, while Latin was at the basis of bronchitis (1814) and tuberculosis (1860), following a tradition of lexical derivation from Classical languages that stretched back for centuries. As for the names of exotic diseases like malaria (1740) and beri beri (1703), it was only possible to import them as loanwords, while attempting to explain their symptoms and the literal meaning of the word(s) in the original language. As for nature, Late Modern times discovered kangaroos (1770) and koala bears (1808), both identified adapting the names heard in Aboriginal languages; eucalypti (1788) and sequoias (1866) – the former was labelled using a Classical prefix and root describing the fact that the blossom was ‘well-covered’, while the latter was named after the Cherokee who devised a writing system for his native language, but had first been labelled Wellingtonia in 1853, giving it the name of the person the botanist wanted to honor and adding the typical Latinate suffix employed for plants. In general, Late Modern vocabulary thus seems to have followed mechanisms of word formation that had been in extensive use for a long time, adapting them to the new requirements of a changing world.
3 Obsolescence and proscription 3.1 Obsolescence Not only did a changing world require new vocabulary, it also required new meanings for existing lexis. According to the OED, between 1700 and 1918 c.3,500 items or uses
55 Late Modern English: Semantics and lexicon were labelled as obsolete (see Table 55.2). If findings are broken down by date as above, this time it is the first half of the 19th century that seems to account for greater change – the indication appears to be that when vocabulary is perceived to be obsolete, it is then time for a new wave of items to enter the lexicon, though of course there may be a certain time gap between actual usage and recorded usage. Table 55.2: Items indicated as obsolete in the OED Time span
Obsolete entries
1700–1750 1751–1800 1801–1850 1851–1900 1901–1918 Total
832 647 1042 980 36 3537
On the other hand, what is recorded in dictionaries can in fact be semantic change, with limited impact on item inventory per se. In such cases the lexical item does not fall out of usage altogether, but loses some of its meanings, thus specializing its range of occurrences; for instance, items that were labelled as obsolete include evolutionist in the sense of an ‘acrobat’ (1833): the noun thus specialized in its scientific sense. Similarly, naturalist lost its meaning of ‘taxidermist’ (1863), to retain only the current one; and binocular no longer meant ‘having two eyes’ (1713), but indicated a process performed by both eyes or indeed an instrument adapted to both eyes: the first reference to a binocular telescope dates from 1738. In other fields, Fenian became much rarer in its folkloric meaning, while developing a specifically political one (1864), while from the language of painting mezzotint became more widespread in engraving as a technique allowing the production of tones and half-tones more accurately. Phrenology in the sense of ‘psychology’ became obsolete, and is nowadays understood to mean what was previously called craniology (1806). In the OED no examples of genius in the sense of ‘inclination, turn of mind’ are recorded beyond 1804, and compliments of condolence or of congratulation were expressed as such in the 18th century, but not later. Indeed, if the focus of this contribution were to extend from lexical items to phraseology, equally interesting developments could be observed; in the case of correspondence, for instance, remarkable change is witnessed in opening and closing salutations (see Nevalainen and Tanskanen 2007). Although on the surface Late Modern, and especially 19th-century, English appears to be reassuringly similar to present-day standard varieties, it is in fact much more complex and divergent than it looks at first sight.
3.2 Proscription lists In addition to natural processes of linguistic obsolescence, due to changes in usage, meaning, and the need to describe new realities while abandoning older, outdated ones, lexical change in Late Modern times also underwent considerable influence from the proscriptive trends visible in the many grammars, dictionaries, and textbooks that stigmatized supposedly “provincial” usage. Lists of proscribed Scotticisms, for
891
892
VI Late Modern English instance, were very widespread in the second half of the 18th century and throughout the 19th century, in line with a more general prescriptive approach to grammar and phonology (see Finegan, Chapter 60). Lexical items, the geographical distribution of which was restricted, were seen to be inadequate for a language that aimed to be “polite”, as they were associated with uneducated usage, and therefore with lowerclass speakers. The fact that some of these same items had probably been in use in literature for many decades, if not centuries, was irrelevant, as the literary texts in which they occurred did not belong to the new “British” canon that was being established (see Crawford 1992), and in which there was no room for geographical specificities – hence the short circuit in linguistic definition in which ‘geographically restricted = provincial = uneducated = lower-class = unacceptable’. The development of 18th-century prescriptivism in a more and more socially evaluative sense is actually quite obvious in many 19th-century grammars, which also comment on lexical choices. In fact, the most striking feature of the earliest booklets is their lack of systematic content organisation, since nouns, verbs, prepositions, adjectives, idioms, and syntactic features all appear in random order, as in the philosopher David Hume’s list of proscribed Scotticisms. This was perhaps one of the very earliest of such lists (see Dossena 2005: 65–72), and was probably meant for private use; however, it got appended to some copies of an early edition of Hume’s (1752) Political Discourses, and since then it circulated widely. Indeed, many of Hume’s examples occurred almost verbatim in 19th-century publications; for instance, in The Vulgarities of Speech Corrected (Anon. 1829) we find more than 200 entries arranged in two columns, “Vulgar Scotch” and “Correct English”, in which as many as 25 items, i.e. 12.5% of the total number, are included with no or minimal variation in the new list, e.g.: p. 234 – p. 236 –
Annual rent of money Interest of money. Dubiety Doubt.
Also in Scotticisms Corrected (Anon. 1855), which includes 565 entries, we find the following items, together with many others that had also been included in Hume’s list: 3. 18. 22.
Has the tailor brought my big coat? say, great coat. He has handed him over the superplus: say, the surplus, or overplus. He showed me great discretion: say, civility.
Although in this case the percentage of entries taken from Hume’s list is lower (7.4%, i.e. 42 items out of 565), it is still interesting to see that they are reported in a very similar way. More than a century later, certain uses were still stigmatized – perhaps, ironically, an indicator of the ineffectiveness of earlier attempts to impose their avoidance. In later works, however, entries are often expanded into sample sentences, or even into anecdotes, in order to make them more memorable, given the explicitly didactic aim of the works. For instance, Mackie (1881: vii) presents his text as “intended as a school book, and, […] as a book of reference in the home”. In it, contents are arranged according to the type of error: incorrect grammar (as in: Were you ringing? instead of ‘Did you ring?’); pleonasm (as in: No passage down this way); impropriety in the use of single words (as in: I do not mind that I ever saw you before); and “violation of idiom, where words individually correct are given in un-English combinations” (1881: 3). The
55 Late Modern English: Semantics and lexicon author does admit that Scots “may be pithy, […] but the usage of polite society in England holds supreme sway in Literature, and we must follow the fashion of the time” (1881: 3). Even though the choice between Scots butter and bread and English bread and butter is “a matter of convention […][,] still the Scotch must yield to English usage. […] So, ‘who do you sit under?’ might easily be justified as figurative language, but being without the stamp of English authority, it must give way to ‘whose church do you attend?’ ” (1881: 3–4). While one form exists in Scots and is indeed found in ordinary, not necessarily uneducated, usage, the fact that it does not correspond to what is perceived to be the more prestigious variety makes it irretrievably unsuitable for adoption into more general (and geographically unrestricted) registers. Indeed, the appeal to “English authority” as the benchmark for linguistic evaluation had been very widespread for at least a century. As we will see in Section 5, this was also the case in Johnson’s Dictionary. Alongside this, another trend focused on etymology and the origin of words for the identification of what were supposedly the most authentic sources of recommended vocabulary. “Scotticism” as a proscribing label was then borrowed to create similar items that conveyed negative evaluations of geographically-restricted features: most famously, in 1781 John Witherspoon (himself a graduate of the University of Edinburgh) formed his “Americanism” on this basis: Americanisms, by which I understand an use of phrases or terms, or a construction of sentences, even among persons of rank and education, different from the use of the same terms or phrases, or the construction of similar sentences, in Great Britain. The word Americanism, which I have coined for the purpose, is exactly similar in its formation and signification to the word Scotticism (1781 Pennsylvania Jrnl. 9 May 1/2; OED, s.v. Americanism).
Interestingly, Witherspoon refers to specificities of usage that affected speakers of all ranks and education levels. Far from being markers of low-class or uneducated usage, both Scotticisms and Americanisms could thus prove tokens of diatopic variation even in what we would nowadays call “standard” varieties. This allowed them to be employed as identity markers when language could function as an emblem of a nation’s identity and individual stance in opposition to Britain (see Section 5). However, the negative connotations of such -isms were difficult to dispel – to the point that later labels, modelled in the same way, such as Australianism, or Australasianism, introduced in the last decade of the 19th century, are recorded among statements expressing serious concern for the “coming degradation of the English language” (OED, s.v. Australianism n.) – a topic that is still of great interest for the general public even to this day (see Crowley, Chapter 61). Like in the case of Scotland, also North American usage attracted the criticism of English observers, whether visitors, like Charles Dickens, or linguistic commentators; texts including “stage Americanisms”, supposed malapropisms, and outlandish turns of phrases were published with the same frequency and success as those that narrated amusing episodes and reported supposedly quaint linguistic uses in “North Britain”. A view of standard usage thus reinforced itself around southern English models, to the detriment of other varieties, both in Britain and overseas. In the United States this attitude provoked a much earlier reaction of linguistic patriotism, most famously represented in the works of Noah Webster (see Section 5), while in Britain it would not be
893
894
VI Late Modern English until the 20th century that fully-fledged vindications of local varieties could develop (for instance with Hugh MacDiarmid in Scotland in the 1920s [see Riach 1991]). The only exceptions to a widespread acceptance of southern linguistic models are recorded in the works of Alexander Geddes (1792) and James Adams (1799): both recommended the adoption of Scots lexical items into English on account of their greater proximity to “the Saxon original” (Adams 1799: 148). Although this appeal to antiquity and purity was to have many followers in the ensuing decades, it was not to have an equally important impact on norm selection on the part of users.
4 Tracing the roots of English The attempt to “streamline” vocabulary, pruning out the forms that were perceived to be merely dialectal (with all the evaluative overtones carried by this label), was accompanied by a similarly important attempt to identify its original roots, so as to reconstruct (and, ideally, maintain) its “original purity”. It is basically for this reason that in Late Modern times a new approach to etymology also began to develop. Starting from an antiquarian interest rooted in the second half of the 17th century, scholars began to investigate place names and other lexical items in an attempt to trace the origins of English beyond the French influence that the Norman Conquest had imposed. This led to a far from culturally and politically neutral kind of “Anglo-Saxonism”, in which the external history of the language was used to explain and defend or criticize lexical change. In the 18th century even Samuel Johnson, in the Preface to his Dictionary, complained that “Our language, for almost a century, has, by the concurrence of many causes, been gradually departing from its original Teutonick character, and deviating towards a Gallick structure and phraseology” (http://andromeda.rutgers.edu/~jlynch/ Texts/preface.html, last accessed September 2011; see also Kolb and DeMaria 2005). However, this comment does not seem to have been taken up by Johnson’s 19thcentury commentators, many of whom preferred to concentrate on his supposedly anti-Scottish definitions (Dossena 2006). On the other hand, the relationship existing between Scots and English was to provide crucial arguments. In the late 17th century, the pamphlet Ravillac Redivivus (Anon. [Hickes] 1678: 77) highlighted the greater proximity of Scots and Northern English to original Saxon forms, comparing it with the distance that Southern English had developed from them on account of Norman influences. As for the controversy on the origins of Scots and English, the debate raged for decades (see Dossena 2006). Some commentators saw the roots of Scots in Pictish: this, in turn, was supposed to derive from Gothic. One of the champions of these views was John Jamieson, who was to become the first important lexicographer of the Scots language (see Section 5), and presented his theory on the origin of Scots in a very extensive study of etymologies of names and place names in the “Introduction” to his Etymological Dictionary of the Scottish Language ( Jamieson 1808/1840: xii–xiii; see also Kidd 1993: 251). As for diverging views, Kidd (1995: 47) stresses that “Interest in things Gaelic [especially following the controversies on Ossian] was more than counterbalanced in the mainstream Scottish political culture of the 19th century by the rise of a Lowland Teutonist identity”; in addition, the negative influence of French was constantly emphasized. The persistence of the antiquarian fashion into the 19th century was also reflected in the continuing search for “pure Saxon” in linguistic matters (see Dury 1992; Milroy 1996). As late as 1888, when Charles Mackay published
55 Late Modern English: Semantics and lexicon A Dictionary of Lowland Scotch with an introductory chapter on the poetry, humour, and literary history of the Scottish language and an appendix of Scottish Proverbs, emphasis was still placed on the closer connection of Scots with the older “AngloTeutonic” vocabulary described as obsolete in English, but still fully comprehensible in Scotland (Mackay 1888: xii). Interestingly, the wish to preserve Anglo-Saxon forms was also part of the agenda in Noah Webster’s linguistic initiatives. In addition, Webster’s fame also relies on his patriotic intentions when launching his spelling reforms, though some of his most extreme views were later repealed. As a matter of fact, his reintroduction of spellings in instead of in words like honour and colour (unlike the ones with instead of in words like centre and theatre) were both more accurate from the etymological point of view, and also found in Scottish English. This, however, did not stop English prescriptivists from stigmatizing his choices. Several publications in Britain attacked specifically American uses in very forceful terms, and it was only in 1902 that British publications began to remark on the different usage of lexical items without apparently evaluative overtones, as in the case of automobile: “On the Continent of Europe and in the United States the usual expression for [s.c. motor-cars] is ‘automobile’ ” (1902 Encycl. Brit. XXXI. 11/1; OED, s.v. automobile adj. and n.).
5 Landmarks in lexicography Fairly dramatic changes in vocabulary and the study of its roots could not but have an impact on Late Modern lexicography. Moving away from lists of “hard words”, or “inkhorn terms”, that had characterized previous centuries, when monolingual dictionaries had first been produced (see Starnes and Noyes 1991), greater attention was given to usage and etymology. In much less than two centuries the British Isles saw the publication of three major dictionaries: Johnson’s (1755), Jamieson’s (1808) and Murray et al.’s, the first part of which was published in 1884 (Murray et al. 1884–1928). As the role played by these works in the standardization of the English language is discussed in Considine (Chapter 66) my aim here is to provide an overview of the philosophies underlying the plans for these dictionaries, since they informed the choices made in lexical recording and defining. Much has been written about Johnson’s Dictionary and his Plan of 1747 (see, for instance, Reddick 1996). In particular, studies have highlighted the discrepancy between the Plan and the Preface in terms of what the lexicographer meant to achieve and his actual product in the light of its tenets. However, it is perhaps useful to focus on the scholarly approach outlined in the Plan, so that its development in later works may be discussed here. As is well-known, Johnson stressed the need for a dictionary that could be of use to the general public, beyond the considerations of critics and theoretical scholars; this would clearly have an impact on contents: It seems necessary to the completion of a dictionary, designed not merely for criticks, but for popular use, that it should comprise, in some degree, the peculiar words of every profession; that the terms of war and navigation should be inserted, so far as they can be required by readers of travels, and of history; and those of law, merchandise, and mechanical trades, so far as they can be supposed useful in the occurrences of common life (http:// andromeda.rutgers.edu/~jlynch/Texts/plan.html; last accessed September 2011).
895
896
VI Late Modern English The Dictionary also set out to stabilize orthography, because “All change is of itself an evil, which ought not to be hazarded but for evident advantage; and as inconstancy is in every case a mark of weakness, it will add nothing to the reputation of our tongue”. The next step would be to fix pronunciation, “the stability of which is of great importance to the duration of a language, because the first change will naturally begin by corruptions in the living speech”. As we can see, the comment belongs to the “complaint tradition” (see Crowley, Chapter 61) and has caused Johnson to be associated so firmly with the prescriptive trends of his times. This is equally observed when, finally, the lexicographer concentrates on etymology, the aim of which is to trace “every word to its original, and not admitting, but with great caution, any of which no original can be found”, so that “we shall secure our language from being overrun with cant, from being crowded with low terms, the spawn of folly or affectation, which arise from no just principles of speech, and of which, therefore, no legitimate derivation can be shown” (http:// andromeda.rutgers.edu/~jlynch/Texts/plan.html). By the time the Preface was written, Johnson had reached the conclusion that: every language has its anomalies, which, though inconvenient, and in themselves once unnecessary, must be tolerated among the imperfections of human things, and which require only to be registred; that they may not be increased, and ascertained, that they may not be confounded: but every language has likewise its improprieties and absurdities, which it is the duty of the lexicographer to correct or proscribe (http://andromeda.rutgers. edu/~jlynch/Texts/preface.html).
Alongside a proscribing attitude, Johnson displayed a distinctly protectionist one in his attempt to defend English from uncontrolled borrowings: The words which our authours have introduced by their knowledge of foreign languages, or ignorance of their own, by vanity or wantonness, by compliance with fashion, or lust of innovation, I have registred as they occurred, though commonly only to censure them, and warn others against the folly of naturalizing useless foreigners to the injury of the natives (http://andromeda.rutgers.edu/~jlynch/Texts/preface.html).
As we saw above, French had long been perceived to be particularly nefarious. However, attempting to stop language change was clearly futile, as Johnson himself had to acknowledge in his Preface: “sounds are too volatile and subtile for legal restraints; to enchain syllables, and to lash the wind, are equally the undertakings of pride, unwilling to measure its desires by its strength” (http://andromeda.rutgers.edu/~jlynch/Texts/preface.html). As for British varieties of English, Johnson meant to exclude dialect words, although several Scots ones do appear (see Dossena 2005: 75–81). At the opposite end of the spectrum Jamieson did not focus on southern English, but precisely on the supposedly “provincial” variety that so many prescriptivists revered in literature, but attempted to iron out of usage in everyday registers. At first Jamieson had identified the specificity of Scottish vocabulary in the legal register – the Scottish legal system being unlike the one viable in England and Wales, on account of the arrangements made at the time of the Union of Parliaments in 1707. Indeed hard word dictionaries, especially legal and medical ones, had had a market for many years. However, Jamieson also stated that his work would “serve to mark the difference between words which may be called classical, and others merely colloquial; and between both of these, as far as they are proper, and such as belong to a still lower class, being mere corruptions, cant terms, or puerilities” (Jamieson 1808/1840: ii).
55 Late Modern English: Semantics and lexicon In this way Jamieson places his work at the intersection of normativity and a more general encyclopaedic interest; in addition, he identifies both different registers and social varieties in Scots, thus highlighting the existence of a ‘proper’ standard and of vulgar speech. In both cases this is a real turning point in the history of Scots lexicography. Towards the end of the century Jamieson’s apparent lack of systematicity was criticized by Fleming (1899: vii); however, what is much more interesting is that Fleming compares Jamieson’s dictionary with a newer enterprise: Murray et al.’s (1884–1928) New English Dictionary on Historical Principles (which had reached letter “H”). The latter was found to be “a remarkably good Scotch dictionary” (Fleming 1899: iv) – but the New English Dictionary would subsequently become world famous (and a touchstone of proper usage) as the Oxford English Dictionary. Unlike Jamieson and many others, James Murray did not think Scots was a separate language, and actually identified it with the northern part of the northern English dialect (Aitken 1995/96: 29). A similarly innovative approach was brought to the OED, about which preliminary views are found in two papers delivered to the Philological Society by Richard Chenevix Trench in November 1857. In these papers, published in London in 1860, and now available in the website of the OED itself (see http://www.oed.com/ public/deficiencies; last accessed September 2011), Trench spoke “On some deficiencies in our English Dictionaries”; among these, the first complaint was the fact that in earlier dictionaries obsolete words are incompletely registered; other items can be antedated; the discussion of some is insufficient, while that of others is redundant. Hence the need for a new approach to lexicography, described by Murray himself in his Romanes lectures in 1900 (see http://www.oed.com/public/romanes; last accessed September 2011). In particular, Murray points out that the method followed for the New English Dictionary is quite unlike the one followed by Johnson, in whose work “illustrations of older words are, in too many cases, […] copied from dictionary to dictionary without examination or verification”(48). In the new enterprise examples are “supplied afresh by its army of volunteer Readers”. The conclusion is that with this dictionary, “permeated as it is through and through with the scientific method of the century, Lexicography has for the present reached its supreme development” (49). This reference to ‘the scientific method of the century’ does in fact sound like an icon of its own time: the positivistic trust in science of the 19th century is possibly one of the most significant legacies of modernity, and it may not be an accident that awareness of its importance should emerge in the considerations of a lexicographer whose attention to the “web of words“ (Murray 1995 [1977]) was to become a cornerstone of modern studies of lexis and semantics. The 19th century saw significant developments also in the lexicography of varieties of English outside Britain. As we mentioned in Section 4, Noah Webster is perhaps the most famous lexicographer beyond England and Scotland, to the point that his own surname has been borrowed to indicate metonymically the dictionary that he first compiled, and that has now gained worldwide relevance. Preceded by a speller, a grammar and a reader, published in the first half of the 1780s as the Grammatical Institute of the English Language, A Compendious Dictionary of the English Language was published in New Haven (CT) in 1806. Its expanded edition, An American Dictionary of the English Language, appeared in 1828 (Webster 1828). Canadianisms were first collected and studied in the second half of the 20th century, when the Dictionary of Canadianisms on Historical Principles (Avis et al. 1967) was launched (see Volume 1, Dollinger, Chapter 119). As for Australia, already in 1790 the School of Oriental and
897
898
VI Late Modern English African Studies in London published Dawes’ Vocabulary of the Language of New South Wales in the Neighbourhood of Sydney (see Leitner et al. 2006), but it was not until 1891 that Lentzner’s Glossary of Colonial English was published, this time including “Australian, Anglo-Indian, pidgin English, West Indian, and South African words” (Lentzner 1891: title page). In the intervening decades, several works on place names were issued, as is often the case when new territories are being explored and the readership is very interested in new geographical features, their unfamiliar etymologies, and the light these may shed on a distant culture. Finally, in 1898 Joshua Lake published his Dictionary of Australian Words (Lake 1898) as The Australasian Supplement to Webster’s International Dictionary (the title acquired by the American Dictionary in 1890). Webster’s attempt to overcome the anglocentricity of models originating in Southern England was thus attaining its first results on a global scale, and the recognition of International English was taking its very first steps.
6 Summary The way in which language was recorded in dictionaries and discussed in publications of antiquaries, or indeed in literary prefaces that referred to glossaries appended to the works of numerous authors, such as Allan Ramsay and Robert Burns (see Dossena 2005: Chapter 5), reflects the scholarly and popular attitudes to standard and dialect vocabulary. The attempts to fix, normalize, and stabilize it soon proved futile, but the attention given to its history and its relationship with other languages was to provide increasing opportunities for investigation and research, even to this day. The attempt to monitor language change through the observation of new formations, borrowings, and the obsolescence of outdated items reflected a certain awareness of the complexity of vocabulary, in which variation, whether social or geographical, was closely related to the dynamics of change. Issues pertaining to social perceptions and evaluations also played a key role in the establishment of what could preserve its status as an acceptable form, and what sounded provincial or old-fashioned, and should as a result be let or made to fall out of use. Similarly, dramatic changes in society, industry and science brought about the need for items that could describe the wealth of innovations being witnessed by speakers of all classes – a trend that gathered speed by the decade, and still affects our own times, when it does not seem to be slowing down, if that is at all possible in an increasingly small and dense world.
7 References 7.1 Primary sources Adams, James. 1799. The Pronunciation of the English Language. (English Linguistics 1500-1800, ed. by R.C. Alston, 72.) Menston: Scolar Press, 1974. Anon. [Hickes, George]. 1678. Ravillac Redivivus. […]. London: Printed by Henry Hills. Anon. 1829. The Vulgarities of Speech Corrected: with Elegant Expressions for Provincial and Vulgar English, Scots and Irish; for the use of those who are unacquainted with grammar. London: F.C. Westley. Anon. 1855. Scotticisms Corrected. London: John Farquhar Shaw. Dawes, William. 1790. Vocabulary of the Language of New South Wales in the Neighbourhood of Sydney. London: School of Oriental and African Studies.
55 Late Modern English: Semantics and lexicon Avis, Walter S., Charles Crate, Patrick Drysdale, Douglas Leechman, Matthew H. Scargill, and Charles J. Lovell. 1967. A Dictionary of Canadianisms on Historical Principles. Toronto: Gage. Geddes, Alexander. 1792. Three Scottish Poems, with a previous Dissertation on the Scoto-Saxon Dialect. Archaeologia Scotica 1: 402–468. Jamieson, John. 1808/1840. An Etymological Dictionary of the Scottish Language […] to which is prefixed A Dissertation on the Origin of the Scottish Language. 2nd edn., ed. by J. Johnstone. Edinburgh: Tait. Johnson, Samuel. 1755. A Dictionary of the English Language. 2 vols. London: Printed by W. Strahan. Lake, Joshua. 1898. Dictionary of Australian Words. Springfield, MA: Merriam. Lentzner, Karl. 1891. Colonial English: A Glossary of Australian, Anglo-Indian, Pidgin English, West Indian, and South African Words. London: Kegan Paul, Trench, Tru¨bner & Co. Mackie, Alexander. 1881. Scotticisms Arranged and Corrected. London: Hamilton, Adams and Co. […]. Mackay, Charles. 1888. A Dictionary of Lowland Scotch […]. Edinburgh: Ballantyne Press. Webster, Noah. 1783–1785. A Grammatical Institute of the English Language. Hartford, CT: Hudson & Goodwin, etc. Webster, Noah. 1806. A Compendious Dictionary of the English Language. New Haven, CT: Hudson & Goodwin. Webster, Noah. 1828. An American Dictionary of the English Language. New York: Converse.
7.2 Secondary sources Aitken, A. Jack. 1995/96. James Murray, master of Scots. Review of Scottish Culture 9: 14–34. Algeo, John. 1998. Vocabulary. In: Suzanne Romaine (ed.), The Cambridge History of the English Language. Vol. IV. 1776–1997, 57–93. Cambridge: Cambridge University Press. Bailey, Richard W. 1996. Nineteenth-Century English. Ann Arbor: University of Michigan Press. Beal, Joan C. 2004. English in Modern Times: 1700–1945. London: Arnold. van Bergen, Linda and David Denison. 2007. A corpus of late eighteenth-century prose. In: Joan C. Beal, Karen P. Corrigan, and Hermann L. Moisl (eds.), Creating and Digitizing Language Corpora 2, 228–246. Basingstoke: Palgrave. Corbett, John et al. (compilers). 2010. Corpus of Modern Scottish Writing (1700–1945). http://www. scottishcorpus.ac.uk/cmsw/ Crawford, Robert. 1992. Devolving English Literature. Oxford: Clarendon Press. Davies, Mark. 2010–. Corpus of Historical American English: 400 million words, 1810–2009 (COHA). http://corpus.byu.edu/coha/ Dollinger, Stefan. 2008. New-Dialect Formation in Canada: Evidence from the English Modal Auxiliaries. Amsterdam/Philadelphia: John Benjamins. Dossena, Marina. 2004. Towards a corpus of nineteenth-century Scottish correspondence. Linguistica e Filologia 18: 195–214. Dossena, Marina. 2005. Scotticisms in Grammar and Vocabulary. Edinburgh: John Donald. Dossena, Marina. 2006. “The Cinic Scotomastic”?: Johnson, his commentators, Scots, French, and the story of English. Textus (Special issue on Samuel Johnson’s Dictionary and the Eighteenthcentury World of Words, ed. by Giovanni Iamartino and Robert DeMaria) 19(1): 51–68. Dossena, Marina and Susan M. Fitzmaurice (eds.). 2006. Business and Official Correspondence: Historical Investigations. Bern: Peter Lang. Dossena, Marina and Roger Lass (eds.). 2004. Methods and Data in English Historical Dialectology. Bern: Peter Lang. Dury, Richard. 1992. Saxonism and the preference for “native” vocabulary. In: Nicola Pantaleo (ed.), Aspects of English Diachronic Linguistics, 133–146. Fasano: Schena. Dury, Richard. 2006. A corpus of nineteenth-century business correspondence: Methodology of transcription. In: Dossena and Fitzmaurice (eds.), 193–205.
899
900
VI Late Modern English Fitzmaurice, Susan M. 2004. Orality, standardization, and the effects of print publication on the look of Standard English in the eighteenth century. In: Dossena and Lass (eds.), 351–383. Fitzmaurice, Susan M. 2006. Diplomatic business: Information, power, and persuasion in Late Modern English diplomatic correspondence. In: Dossena and Fitzmaurice (eds.), 77–106. Fleming, J. B. Montgomerie. 1899. Desultory Notes on Jamieson’s Scottish Dictionary. Glasgow/ Edinburgh: William Hodge and Co. Go¨rlach, Manfred. 1999. English in Nineteenth-century England. Cambridge: Cambridge University Press. Go¨rlach, Manfred. 2001. Eighteenth-century English. Heidelberg: Winter. Hume, David. 1752. Scotticisms. In: Political Discourses. Edinburgh: Printed by R. Fleming, for A. Kincaid and A. Donaldson. [unnumbered pages] Kastovsky, Dieter. 2006. Vocabulary. In: Richard Hogg and David Denison (eds.), A History of the English Language, 199–270. Cambridge: Cambridge University Press. Kidd, Colin. 1993. Subverting Scotland’s Past. Scottish Whig Historians and the Creation of an Anglo-British Identity, 1689–c.1830. Cambridge: Cambridge University Press. Kidd, Colin. 1995. Teutonist ethnology and Scottish nationalist inhibition, 1780–1880. Scottish Historical Review 74: 45–68. Kolb, Gwin J. and Robert DeMaria (eds.). 2005. Johnson on the English Language. The Yale Edition of the Works of Samuel Johnson. New Haven: Yale University Press. Kyto¨, Merja, Mats Ryde´n, and Erik Smitterberg (eds.). 2006. Nineteenth-century English. Cambridge: Cambridge University Press. Lass, Roger. 2004. “Ut custodiant litteras”: Editions, corpora and witnesshood. In: Dossena and Lass (eds.), 21–50. Leitner, Gerhard, Clemens Fritz, and Brian Taylor (eds.). 2006. Language in Australia and New Zealand. A Bibliography and Research Database (1788–Present). Berlin/New York: Mouton de Gruyter. Milroy, James. 1996. Linguistic ideology and the Anglo-Saxon lineage of English. In: Juhani Klemola, Merja Kyto¨, and Matti Rissanen (eds.), Speech Past and Present – Studies in English Dialectology in Memory of Ossi Ihalainen, 169–186. Frankfurt: Peter Lang. Mugglestone, Lynda (ed.). 2000. Lexicography and the OED: Pioneers in the Untrodden Forest. Oxford: Oxford University Press. Mugglestone, Lynda. 2006. English in the nineteenth century. In: Lynda Mugglestone (ed.), The Oxford History of English, 274–304. Oxford: Oxford University Press. Murray, James A. H., W. Bradley, A. Craigie and C. T. Onions (eds.) 1884–1928. A New English Dictionary on Historical Principles. Oxford: Oxford University Press. Murray, Katherine M. E. 1995 [1977]. Caught in the Web of Words: James A. H. Murray and the Oxford English Dictionary. New Haven: Yale University Press. Nevalainen, Terttu and Sanna-Kaisa Tanskanen (eds.). 2007. Letter Writing. Amsterdam/Philadelphia: John Benjamins. Reddick, Allen. 1996. The Making of Johnson’s Dictionary 1746–1773. Rev. edn. Cambridge: Cambridge University Press. Riach, Alan. 1991. Hugh MacDiarmid’s Epic Poetry. Edinburgh: Edinburgh University Press. Simpson, John (ed.). 2000–. The Oxford English Dictionary. 3rd edn. online. Oxford University Press. www.oed.com/. Starnes, DeWitt T. and Gertrude E. Noyes. 1991. The English Dictionary from Cawdrey to Johnson, 1604–1755. 2nd edn. Amsterdam/Philadelphia: John Benjamins. Tieken-Boon van Ostade, Ingrid. 2005. Eighteenth-century English letters: In search of the vernacular. Linguistica e Filologia 21: 113–146.
Marina Dossena, Bergamo (Italy)
56 Late Modern English: Pragmatics and discourse
901
56 Late Modern English: Pragmatics and discourse 1 2 3 4 5 6
Introduction Genre-based studies Pragmatic expressions and functions Politeness Future perspectives References
Abstract The evolution of discoursal and pragmatic aspects of English during the Late Modern period is intricately linked to contemporary socioeconomic developments. The spread of literacy, the increasing social and political power of the upper middle classes, and new opportunities for social advancement in a rapidly evolving class system led to a heightened awareness of language use and to new attitudes to language in the first part of the period. Subsequently, a reaction to “refined” language resulted in an appreciation of “real” English. An important development in the English of this period is the divergence and conventionalization of genres. The chapter exemplifies this trend by looking at scientific writing and newspaper language. Research in this area remains fragmented, with a great many small-scale studies undertaken using a wide variety of methods, often on narrow ranges of textual data. The chapter focuses only on English in the British Isles.
1 Introduction 1.1 Background Both the Late Modern period and the sub-disciplines of historical discourse analysis and historical pragmatics have been relatively neglected until the last decade of the 20th century. But a number of research initiatives and methodological developments have combined, with some acceleration in the 1990s and 2000s, to illuminate various pragmatic and discoursal aspects of English after 1700. Researchers have benefited from ever easier access to digitized historical texts from wide-ranging contexts. The tendency in corpus linguistics away from “general language” corpora towards genrespecific corpora has been echoed in corpus-based historical linguistics (see Kyto¨, Volume 2, Chapter 96). The inauguration in 2001 of a conference series on Late Modern English (Dossena and Jones [eds.] 2003; Dossena and Tieken Boon van Ostade [eds.] 2008) helped to raise the profile of the period as a focus of linguistic research, and since 2000 the Journal of Historical Pragmatics has provided a new forum. This section provides some background to the main frameworks and methods adopted by research into pragmatic and discourse aspects of Late Modern English (LModE). Section 2 looks at research on genre development in the Late Modern period and at some discoursal features of particular language genres. Section 3 deals with studies of pragmatic expressions such as discourse markers and of some pragmatic functions. Section 4 sums up the current situation and looks at perspectives for further research. The focus of this chapter is on English in the British Isles (cf. Volume 2, Sections XIV and XV). Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 901–915
902
VI Late Modern English Changes in English discourse practices during the Late Modern period are best understood in the context of the socio-economic and political developments of the time (cf. Beal 2004; Beal, Chapter 5). The 18th and 19th centuries saw unprecedented population movements in Britain as industrialization and urbanization developed. The period also saw a decline in the influence of the aristocracy and the emergence of a powerful and substantial urban middle class. Radical transformations of social class structures and the rapid spread of literacy led to increased language awareness and shaped new attitudes towards language. These attitudes, and the changes in discourse practices now seen as characterizing the period, are very much a product of the rising influence of the upper middle classes in the new, evolving class structure. Few people in the 18th and early 19th centuries could afford books, but widespread literacy created a market for circulating libraries, which from the 1780s onwards helped turn reading into a major form of entertainment. The increased accessibility of print, and the generalization of the print culture needed for standardization to take root, enabled new written genres to establish themselves. The early 18th century had seen the emergence of the so-called “new rhetoric” which created new ideals for written discourse, and prestigious writers and essayists provided models. Oral styles in writing came to be looked down upon in literate circles with social aspirations, where more elaborate and “refined” language, often involving latinate lexis and syntactic subordination, was valued (cf. Adamson 1998). Learning to write often involved learning the emerging standard variety while continuing to use a regional variety when speaking. Spoken language in turn was affected by class aspirations and by the desirability of appearing literate, although it was not until the 19th century that regional spoken language became so strongly stigmatized (cf. Go¨rlach 2001). The material benefits that accrued from using “refined” language created a lively market for prescriptive books on usage, rhetoric, and grammar (Finegan, Chapter 60). Newly-established journals for the middle classes, such as Tatler and Idler, carried articles about language usage (cf. Finegan 1992). Oral or “vulgar” usages were held in contempt. The citing of proverbs, for instance, came to be regarded as vulgar, as characteristic of the illiterate (McIntosh 1998: 10–11). “Oral” and “literate” cultures thus came to be more sharply distinguished. As Hudson (1994: 163) shows, “an important legacy of eighteenthcentury thought is the perception that writing and speech are distinct and dissimilar forms of communication with quite different powers and functions”. The rejection by the rising middle classes of what was perceived as the speech styles of the illiterate was part of a wider rejection of popular culture and traditional arts in general, just as the cultivation of refinement in language was paralleled by elegant manners and elaborate clothing. The prevailing attitude is captured by Lord Chesterfield’s comment in a letter to his son that “style is the dress of thoughts” (cited in Knowles 1997: 137). Plainness of style became disfavored. There followed something of a backlash during the Romantic movement from the end of the 18th century onwards. The revered status of the written word began to decline, as nostalgia set in among the upper middle class for the lost charms of oral society and of the more natural spoken language. Elevated styles were frowned upon, and by the early 20th century Fowler and Fowler’s (1924 [1919]) prescriptive The King’s English was advocating the use of short, familiar, concrete words of Germanic origin rather than longer, more abstract, or Romance ones, chiding those who would say violin for fiddle, for instance. Another reflection of this new ideal is the consciously abundant
56 Late Modern English: Pragmatics and discourse
903
use in some early 20th century literature of phrasal and phrasal prepositional verbs, which had come to be considered colloquial and inferior during the fashion for refinement . These tensions between refined and plain language, and the place of each, interacted in the 19th century with emerging nationalist ideologies and with lasting awareness of the Latinate versus Anglo-Saxon origins of words: a division that, as Fairman (2006: 80) points out, “had complex historical, social and racial […] connotations”.
1.2 Frameworks and methods Pragmatic aspects of the language of the Late Modern period have been analyzed using a number of different theoretical and methodological frameworks; indeed, there has been a healthy focus on discussing methodological issues, as is evidenced, for instance, by Taavitsainen and Jucker (2007), Arnovick (1999: Chapter 1), Fitzmaurice and Taavitsainen (2007), and Jucker et al. (1999). These frameworks can be roughly characterized by parameters such as: a. the size of the linguistic units of interest; b. whether the analysis is qualitative or quantitative; c. whether the study is semasiological (of the pragmatic and/or discoursal functions of a form) or onomasiological (of how a pragmatic or discoursal function is expressed); d. whether the data is taken from a single variety or several; e. from “general language” texts or from specific genres; and f. whether the focus is on the state of the language at a given time (synchronic or “historical pragmatics”) or on language change (“diachronic pragmatics”). Some of these frameworks and their main characteristics are shown in Table 56.1, with examples. Table 56.1: Some frameworks used for historical-pragmatic and discoursal studies of Late Modern English Approach
Framework
Characteristics
Example
pragmatic and language-based approaches
neo-Gricean
qualitative or quantitative; semasiological; often general, standard language; word- or expression-level; diachronic quantitative (corpus-based, multi-variate analysis); variationist; text-level; synchronic qualitative; onomasiological; expression-level; synchronic
Brinton 2007
social network analysis
qualitative; variationist; textlevel; panchronic
Fitzmaurice 2007
genre analysis and rhetorics literary and culturalstudies approaches
qualitative; onomasiological; variationist; text-level qualitative; textual-history; synchronic and diachronic
Bazerman 1988 McIntosh 1998
multidimensional texttype analysis politeness theory
discourse-analytic and context- based approaches
Biber and Finegan 1989 Bijkerk 2004
904
VI Late Modern English Table 56.1 represents a broad interpretation of the term “historical pragmatics”; others may prefer a narrower definition. Some authors have treated historical pragmatics and historical discourse analysis as almost interchangeable terms (Jucker 2006: 329; see also Traugott 2003: 539 and Brinton 2001a: 139–140). Yet pragmatics, as the study of context-bound aspects of language, has different goals and methods from discourse analysis, which is qualitative, variationist, and more literary in approach. Studies of changing discourse practices overlap with the larger body of sociolinguistic work on the period (see Smitterberg, Chapter 59). The main centre of interest for pragmaticists focussing on the period has been to trace innovations in the area of highly context-bound “pragmatic” expressions such as discourse markers and idioms, discourse connectives, politeness markers, terms of address, and so on. In studies of linguistic change, the more pragmatic approach tends to focus on internal language change, while research starting from a sociohistorical context naturally emphasizes external language change. However, the approaches mentioned in Table 56.1 are by no means mutually exclusive and most studies involve more than one of these perspectives. The purpose of many studies is, of course, the accumulation of evidence in support of a theoretical position, for which the period may be incidental. Others, however, focus on the description, documentation, and comparison of the language and texts of the period. Given the wide variety of both the language data and the research aims, it is perhaps useful to separate research that focuses on the language and genres of the LModE period from research that focuses on an expression type or a pragmatic function, often in order to support a theory of language change. This division broadly reflects the distinction between “discourse-oriented historical linguistics” (Brinton 2001a: 140) on one hand and synchronic or diachronic discourse analysis (Brinton 2001a: 139–140) or “historical pragmatics” (Jucker 2006) on the other.
2 Genre-based studies 2.1 Genre evolution As pointed out by Bazerman (1988: 62–63), “the formation of a genre reveals the forces to which textual features respond”, and the new form of communication is in turn “powerful enough to influence other forms of communication and the social structure of the community which uses it”. Studies of the evolution of genres during the LModE period reveal accelerating diversification and specialization, in parallel with increasing professionalization and knowledge specialization. There is a temptation to categorize historical texts using present-day genre labels (based on the content, the distribution, and the language of texts) that can be anachronistic. Until the late 17th century, it was more natural to distinguish among styles, a notion inherited from Roman rhetoric. Texts written in the middle style, for instance, might display similar linguistic features and discourse conventions across different topics and places, but would be clearly distinguished by writers and readers from texts in the plain style. The emergence of diverse genres and the dwindling importance of the “styles” is an important development of the 18th and 19th centuries. Up to half of the texts appearing in the Philosophical Transactions of the Royal Society of London through the 18th century were in the form of letters (48% in 1775 according to
56 Late Modern English: Pragmatics and discourse Atkinson 2001: 49). Should genre-based studies include them with “letters” or with “scientific articles”? They are not so much examples of “letters” or of “scientific articles” as products of a print-enabled public sphere of gentlemanly culture, where ideas were discussed, opinions formed, and social and political power negotiated. The polite letter belonged to this sphere; it was “an emblem of genteel culture” (Atkinson 2001: 63). At the same time, there was growing awareness of contemporary sub-languages: Campbell writes in 1776 of “professional dialects” existing for the conduct of commerce or medicine, etc. (cited in Finegan 1998: 553). And in the 19th century such text types increasingly diverged (Go¨rlach 2001: 208–215), paving the way for a multiplicity of new genres and sub-genres. Comparing English texts from the 17th to the 20th centuries, Biber and Finegan (1989) find that the range of different text types (defined by linguistic, not situational, features) is greatest in the 20th century. Present-day genres that started to take off in the 18th century, due to the rapid increase in literacy, include personal correspondence, essays, newspaper and journal articles, advertisements, and scientific articles among others. The gradual adoption of conventional styles of discourse by numerous discourse communities has been at least partially documented for these areas. The rest of this section looks at the examples of scientific discourse and newspaper and journal language, and finishes with a look at cross-genre comparison.
2.2 Scientific and medical writing English-language scientific discourse was still relatively new at the beginning of the Late Modern period, and use of Latin had by no means yet died out. Nevertheless, from the early issues in the late 17th century of the Philosophical Transactions of the Royal Society of London a distinctive rhetorical style began to emerge, together with a consciousness of language, exemplified by Boyle’s (d. 1691) advocacy of a “rhetoric of immediate experience” (Atkinson 1992: 339; see also Biber and Finegan 1989: 512–513) to recount concrete events and observations. Using rhetorical analysis and the “tools of literary criticism”, and drawing on data from The Philosophical Transactions of the Royal Society, Bazerman (1988) shows how the form of the experimental article slowly took shape towards the end of the 18th century. Both article content and article form changed in response to new values of the evolving discourse community, which became gradually less interested in the particular observation or experiment and more interested in general theories and laws of nature. The term “experiment” became more clearly differentiated from observation, its sense narrowing to refer to activities undertaken to test hypotheses and provide proofs. As replicability became recognized as crucial to proof, so descriptions and justifications of method became longer and more detailed, to enable replication and the resolution of differing results. The reporting of the results themselves also became detailed and quantitative. The focus of interest having shifted from the experimental event to the conclusions and proofs, it became common to report a series of experiments pointing to a single conclusion. The experiment-result ordering in articles gave way to hypothesis-experiment ordering, and articles became more likely to start with a discussion of the general problem area. These developments resulted in ever-longer articles. The discourse organization thus reflected the changed role of the experiment, from an event worthy of description in itself to a means of resolving a question and defending a theory. Whereas earlier experimental reports were “simply a matter of news”
905
906
VI Late Modern English (Bazerman 1988: 77), by 1800 they had developed into a recognizable genre shaped by the need to persuade in accordance with the values of empiricism of the scientific community. The history of the scientific article since the 17th century has been traced in a series of publications by Atkinson. Medical texts are addressed by Atkinson (1992), who analyzes samples of writing taken from seven years across the 1735–1985 history of the Edinburgh Medical Journal. His findings reveal the evolution of the modern medical research article via changes in content, authorial presence, and discourse structure (Atkinson 1992: 343–348). Compared with later ones, the earlier articles, mostly written as letters, contain more references to the authorial first person, more active verbs, and more detailed descriptions of the events witnessed at first hand by the author. The sequential narratives of earlier articles, in which one case is recounted, give way first to multiple cases, then to the integration of cases into more abstract discussions, and finally, by the mid 20th century, to the convention of introduction-method-result-discussion still in use today. The medical content involves increasing abstraction over cases: from the 18th century individual case, patient-by-patient model, via the adoption in the late 19th century of the laboratory science model, to the almost complete elimination of case material from medical research writing in the second half of the 20th century. Atkinson goes on to measure 70 of the texts against Biber’s (1988) text type dimensions (see Section 2.4), showing that the greatest change over the period was a decrease in authorial involvement and a corresponding increase in informativeness. The articles also became gradually less narrative and less overtly persuasive. These findings are very similar to Bazerman’s (1988) findings on scientific research articles. Atkinson (1999, 2001) applies the same method of analysis to texts of the Philosophical Transactions of the Royal Society of London between 1675 and 1975. Scientific articles are found to change over the period from neutral on Biber’s (1988) Informational-Involved dimension to very informational. This is largely accounted for by the increase in noun density and noun compounds at the expense of involved features such as first-person pronouns and private verbs. On the narrative dimension, articles evolve from almost neutral to become much less narrative, and (though the results are less linear here) they become more abstract, as implied by the higher density of passive forms. By contrast, Oldireva Gustafsson (2006) finds no significant increase in the use of passive forms in 100,000 words of 19th century scientific monographs. Atkinson (2001: 61) concludes that “the author-centred rhetoric was gradually replaced by more object-centred norms”. In her qualitative study of natural history and medicine articles from the Philosophical Transactions of 1765–1768, Valle (2004) also finds abundant authorial evaluation, especially attributive adjectives evaluating both people and the objects and phenomena reported on, reinforcing a strong authorial presence. A decline in the occurrence of attitudinal and modal features in scientific research articles over 1650–1990 is reported by Biber (2004: 125) (see Section 2.4 below). Compatible findings are reported by Salager-Meyer and Zambrano (2001), working on the language of medicine. In their quantitative comparison of English and French medical writing in the 19th and 20th centuries, they look at the ways in which the texts encode conflict and polemic. Scientists writing in English are found to diverge from their French-writing counterparts by using an increasingly more indirect, impersonal style. These and other studies all point to the gradual reduction in authorial presence and increase in informativeness and objectivity in scientific writing. The conventional
56 Late Modern English: Pragmatics and discourse experimental report that started to form during the 18th century persisted well into the 19th century, and the trend towards giving a greater place to theory at the expense of descriptions of method continued during the 20th century.
2.3 Newspapers and journals Publication of daily newspapers in London dates from the beginning of the 18th century, and their subsequent history has attracted much scholarship (see Fries, Chapter 67). The period between the late 18th century and the early 20th century witnessed the rise and rise of the newspaper. By 1800, it is estimated that perhaps a third of the population of London, or 250,000 people, habitually read a newspaper. Most provincial towns published their own paper. Copies were passed around and were read aloud in homes and ale-houses for those unable to read themselves (Barker 2002). The 18th century also saw a surge in the numbers of journals being published in the British Isles. Overall trends that have been documented by research on newspaper language include a lowering of the level of formality, diversification into sub-genres, and changes in information structure, notably increased information density. Finegan (1992) discusses how very varied were the styles to be found in new journals of the early 17th century. Individual writers used both plain and elaborate styles at different times (Finegan 1992: 110), in journals that circulated in relatively small and homogeneous communities. Conventions for the medium were yet to be established. As literacy was acquired by more and more classes through the 19th century, so news audiences grew, and so the balance of content, as more and more new papers became established, began to shift away from serious news and towards gossip and other entertaining trivia (Aitchison 2007: 84–92). Sub-genres of newspaper and journal language soon emerged, including, of course, the news report, which came into its own after the invention of the telegraph in 1844. The use of traditional chronological narrative form for news stories gradually gave way to the news-specific “inverted pyramid” text, whereby events are recounted in order of decreasing importance, and all the essential elements are introduced in the first paragraph. Information is densely packed, largely through dense NP structures containing noun modifiers. In a quantitative study of the NN structure from the 17th century to the 20th, Biber (2003) shows that in newspaper language, the huge increase in frequency of NN in the last hundred years is largely accounted for by the increase in N + common N. He suggests two possible reasons: first, new technologies have made repeated editing easier, and second, the “informational explosion” has created pressures to be succinct. Bell (2003) illustrates the radical changes to news discourse structure over the 20th century, with an analysis of the 1913 reporting of Scott’s 1912 South Pole expedition. It shows, compared with the present-day news report genre, more and shorter articles, more and longer headlines, and less information compression. Another sub-genre of newspaper and journal language that underwent radical changes in the Late Modern period is advertisements. By comparing advertisements across the 18th and 19th centuries, including those in Defoe’s Review of 1704–13 with those in The Newspaper of 1844, Go¨rlach (2002) charts the divergence of advertising discourse from journal article discourse. Advertisements were little different from other descriptive texts in the 18th century, but during the 19th century became distinctive by their typography, their shortness, their placement, and their increasing use of
907
908
VI Late Modern English formulaic expressions such as most respectfully entreats the public to … and literary devices such as rhymes and quotations. This is another example of divergence, of conventionalization of a distinctive group of discourse features. Gieszinger’s (2000) study of advertisements in The Times over the 19th and 20th centuries shows similar changes. In the late 18th and early 19th centuries, advertisements are primarily informative and visually similar to the rest of the print. From the end of the 19th century they become more conspicuous and more conventionally structured and therefore recognizable. Over the past three centuries, then, and notwithstanding much intra-genre variation, newspaper and journal language, designed for an increasingly wide and ultimately universal readership, has evolved into myriad sub-genres. Scientific articles, designed for an increasingly professionalized community, have become highly conventionalized and more abstract and impersonal.
2.4 Cross-genre comparison Brinton (2001a: 151–152) points out that “studies of changes in discourse or genre have focused almost exclusively on changes that result from the shift from the oral to the written medium” and continues: “One might […] question whether the focus on oral and written features […] is the most useful one”. The focus does now seem to be changing. The findings from genre-based studies such as those mentioned above suggest that the medium (oral or written) is much less relevant as a parameter shaping discourse styles than distance (geographical, social, or temporal), milieu (social network practices), or contemporary beliefs about language. Trends such as those mentioned in Section 2.3 in the context of science writing or newspaper language have been shown to stem from the values of the discourse communities involved. Ultimately it is the evolving values and practices of the relevant networks and communities that shape discourses. While pragmaticists and discourse analysts have focused mainly on particular genres, authors, or texts, some advances have been made in multi-genre comparison. In particular, the multidimensional method mentioned in Section 2.2, developed by Biber and colleagues, has been applied to genre-diversified historical corpora. The “dimensions” result from a cluster analysis of frequency counts of typically 60–70 different linguistic features. The resulting clusters of features characterizing groups of texts are interpreted as textual dimensions, such as Informational-Involved and Narrative-Non-narrative. The method is rather a blunt tool when it comes to diachronic studies, unless the choice of forms is based on each period concerned and unless account is taken of semantic shifts and changes in usage for the forms that recur across periods. Nonetheless, this approach has produced interesting results that are in line with the findings of qualitative and literary studies of the 17th to 20th century period. Biber and Finegan (1989) compare the three genres of fiction, essays, and letters over the 17th to 19th centuries and discuss two main findings. First, more variation is found overall within each genre in the 18th century than in either of the other two centuries. This variation narrows down to more constraining genre norms in the 19th century (Biber and Finegan 1989: 507). Second, Biber and Finegan claim that the 18th century shows more “elaborated” (context-independent) language than either the 17th century or the 19th century. On the Informational-Involved dimension, they note a swing in the 18th century
56 Late Modern English: Pragmatics and discourse towards a less involved style, and back towards more involvement in the 19th century, reflecting the prestige attached to elaborate, formal, polite language that reached a peak in the 18th century. Smitterberg (2005) applies Biber’s multidimensional analysis method to A Corpus of Nineteenth-century English (CONCE) (see Kyto¨ et al. 2000) to show the frequency and distribution of the progressive aspect as it rapidly increases over the course of the century. These findings likewise reveal important differences in frequency among different genres, consistent with the divergence found in other studies. Geisler (2002) uses a multidimensional approach (with a subset of Biber’s 1988 features) on 187 text samples from the CONCE corpus to investigate seven genres across the 19th century: Drama, Trials, Letters, Fiction, Parliamentary Debates, Science, and History. Clear differences are found among the seven genres. In particular, there is a division between expository text types found in History, Debates, and Science, and non-expository types, found in Drama, Fiction, Letters, and Trials. What is less clear is whether there is significant change over the century.
3 Pragmatic expressions and functions 3.1 Discourse markers and connectives Discourse markers are often informal and subjective and may be confined to particular social circles. Their much greater frequency in oral language makes them more difficult to study in diachrony where only written data is available. Since the life cycle of a discourse marker is quite short, they provide a good area for language change research into reanalysis and semantic and pragmatic change. They reveal, for instance, how certain types of expression lend themselves to increasingly attitudinal and interpersonal uses, with resulting subjectification of sense. Discourse markers are known for their frequent renewal. Particularly subject to sociolinguistic factors and fashion, they tend to be “caught” easily, spreading quickly among social networks. Choice of markers therefore can reflect age, social position, and so on. Discourse markers date quickly: many of the most frequent discourse markers and connectives of the 20th century arose only in the 18th or 19th century, including of course, after all, still, I say. Brinton (2007, 2008) exemplifies discourse-marking research that reveals important regularities in language change. Brinton examines the origins and development of expressions I say, you know, what’s more, and which is more and shows how they were reanalyzed into discourse markers. The evolution of the pragmatic import and discourse functions of the expressions is traced through numerous examples. Brinton (2001b) deals with the development of interpersonal pragmatic markers, showing how, through usage, specific pragmatic functions developed out of a verbal predicate. There is a theoretical point to these studies, which use the diachronic data as evidence for a theory of language change – here, the changes are used to support a view of grammaticalization. Along similar lines, Finell (1992) examines the use in personal letters of a number of discourse markers that developed into topic changers in the 18th and 19th centuries, such as still, anyhow, of course, and however. And Lewis (2003) traces the grammaticalization of the EModE prepositional phrase of course through the 18th and 19th centuries to the increasingly polysemous discourse marker of the 20th century.
909
910
VI Late Modern English Tag questions have a similar pragmatic role to discourse markers, and are more frequent in spoken English than in written. Using the texts of plays to approximate spoken usage, Hoffmann (2006) shows a very significant increase in the frequency of tag questions from less than 100 per million words in the second half of the 18th century to around 430 per million words after 1900. The comedy genre displays close to a sixfold increase over the 19th and 20th centuries. The relative proportions of different question tags are found to vary little throughout the period, but an increase in the proportion of “constant tags” (positive verb + positive tag) in the 18th century is reversed again after the middle of the 19th century.
3.2 Speech acts and onomasiological studies Much of the research mentioned so far, including the corpus-based studies, relies heavily on the semasiological approach: it involves looking in one way or another at the frequency and distribution of chosen linguistic forms. The complementary (and methodologically more difficult) approach, whereby the researcher sets off from a functional area to investigate how it is linguistically encoded, is well exemplified by the collection of papers in Jucker and Taavitsainen (2008). An interesting example that mixes the two approaches is Arnovick (1999), who studies the evolution of the expression of particular speech acts such as promising, blessing, greeting, swearing, and so on. Also adopting a mixed approach, Biber (2004) uses a genre-diversified historical corpus to examine the expression, the frequency and the distribution of one area of meaning – “the entire system of stance devices” (Biber 2004: 107) – in the period 1650–1990 in Drama, Letters, Newspaper Reportage, and Medical Prose. One significant finding is the apparently increasing register divergence: Drama and Personal Letters show increase in the frequency of stance markers whereas in medical prose there has been a marked fall (see Section 2.2 above). The study shows how modal verbs have declined over the three centuries, while stance adverbials and semi-modals have increased in frequency. According to Biber, these findings suggest that “stance meanings have come to be expressed to an increasing extent across these historical periods” (Biber 2004: 126) with the most rapid change occurring in the 20th century. Given that the frequencies are proportional, some other notional area(s) must have decreased in frequency. Most of the increase in stance marking is accounted for by epistemic markers, not affective or attitudinal ones; and, as Biber points out, the notion of stance and its identification in linguistic features needs refining. There are both theoretical and methodological problems with identifying speech acts or pragmatic functions and their linguistic realizations. Nevertheless, these types of study are preparing the way for larger-scale, corpus-based investigations that have the potential to provide some background against which the more isolated research findings can be put into perspective and more richly interpreted.
4 Politeness Wide-ranging socioeconomic and political changes, including new class structures and new distributions of power benefiting the emergent bourgeoisie, led to increased opportunities for social mobility. The spread of what came to be known in the later 19th century as public schools (for the sons of the wealthy), including the incorporation of many
56 Late Modern English: Pragmatics and discourse grammar schools into the system, favored the creation of new social networks beyond family and neighbourhood. There quickly developed a flourishing market for etiquette guidance of all kinds, including “courtesy books” and an array of prescriptive language and grammar books. The plethora of such publications at the end of the 18th century and beginning of the 19th century suggests acute social awareness, including a heightened sensitivity to the role of language in self-revelation. This sensitivity is also apparent in the careful attention paid to appropriate, polite and often intricate interpersonal expression in contemporary discourse. At the beginning of the 18th century, politeness emerged as a “powerful cultural ideal” (McIntosh 1998: ix), expressed through language and manners among other means. An increasingly common approach adopted in diachronic studies of politeness is that of social network analysis, following work by Fitzmaurice on well-known 18th century London literary discourse communities (see Fitzmaurice 2007). At the same time personal correspondence has been the focus of much recent research into the discourse and pragmatics of the period (see Dossena and Tieken-Boon van Ostade [eds.] 2008; Palander-Collin and Nevala 2010). Bijkerk (2004) adopts a social network approach to investigate how the letter-closing formulae yours sincerely and yours affectionately may have come to replace the older your most humble/obedient servant, and which types of people seem to have led the change. Work on address terms by Nevala (2007, 2009) focuses on the expression of social identities and social relationships within a social speech community through analysis of referential expressions in letters and journals. It is the vocabulary of manners that is the subject of Orduna Nocito’s (2007) study. Through contemporary dictionary definitions and usage in the Spectator journal of the early 18th century, she examines a range of terms relating to manners and charts some semantic drift. Tieken-Boon van Ostade and Faya Cerqueiro (2007) is a diachronic pragmatic study of the development of please. As well as proposing a developmental path for PDE please from be pleased to, the authors describe the distribution in some 18th century correspondence of older forms if you please and pray. It is a good example of research that includes both study of structural reanalysis and consideration of the social and pragmatic factors propelling the change. Watts (2002) considers the relationship between polite language and the emergent standard language, suggesting that the standard should be seen as a development of the concern for politeness that is so characteristic of the LModE period.
5 Future perspectives Increasing interest in the role of discoursal and pragmatic factors in internal language change has focused attention on methods for historical and diachronic discourse analysis and pragmatics in the absence of native speakers and spoken data. Yet interest in the pragmatics and discourse of the Late Modern period has been slow to develop, and the field remains fragmented, with findings often to be sought at the margins of work centered on other periods and linguistic topics. Discourse analyses of the language of the 18th century clearly reflect the state of flux in the written language of that time, as its status and functions radically evolved. Studies such as those mentioned above almost all point to the Late Modern period being one where genres, situationally defined, became more numerous, more differentiated from
911
912
VI Late Modern English one another and more internally consistent. Much of the research also suggests gradually increasing expression of speaker/writer attitude or involvement as revealed by frequencies of “pragmatic” expressions. More work is needed, however, both to resolve theoretical questions of what is meant by “pragmatic” in this context, to replicate studies on more and more textual data in order to empirically confirm findings, and to investigate further the changes that may have occurred over the period in the implicatures generated by linguistic expressions, constructions, and discourse patterns. Future work on the micro and macro levels will allow a more comprehensive view to be built up of patterns of usage of pragmatic features and of developments in discourse conventions in the LModE period. Inspired by social network theory, research into the circulation of discourse within particular and well-documented social networks of the period reveals how pragmatics can drive language change, and how specific changes or expressions propagate through a network and beyond it (e.g. Fitzmaurice 2007). At the other extreme, very large text bases are needed for the kind of statistical analysis of frequencies and distributions that allows long-term linguistic changes to be traced. Equally, large amounts of text are needed for synchronic characterizations of period genres and varieties. The results of descriptive studies based on limited text samples or on few authors often cannot be extrapolated beyond that database. One stumbling block has been the quality and availability of electronic corpora: while good representative corpora of present-day English and of Old, Middle and Early Modern English have become widely available, studies of the English of the 19th and early 20th centuries have suffered from sparse data and insuperable copyright problems. New, larger, genre-specific and genre-diversified corpora representative of both standard and nonstandard varieties of English will greatly benefit research in this area and increase the reliability of findings.
6 References Adamson, Sylvia. 1998. Literary language. In: Romaine (ed.), 589–692. Aitchison, Jean. 2007. The Word Weavers: Newshounds and Wordsmiths. Cambridge: Cambridge University Press. Aitchison, Jean and Diana M. Lewis (eds.). 2003. New Media Language. London: Routledge Arnovick, Leslie K. 1999. Diachronic Pragmatics. Amsterdam/Philadelphia: John Benjamins. Atkinson, Dwight. 1992. The evolution of medical research writing from 1735 to 1985: The case of the Edinburgh Medical Journal. Applied Linguistics 13(4): 337–374. Atkinson, Dwight. 1999. Scientific Discourse in Sociohistorical Context. The Philosophical Transactions of the Royal Society of London, 1675–1975. Hillsdale, NJ: Lawrence Erlbaum. Atkinson, Dwight. 2001. Scientific discourse across history: A combined multi-dimensional/rhetorical analysis of the Philosophical Transactions of the Royal Society of London. In: Susan Conrad and Douglas Biber (eds.), Variation in English: Multi-Dimensional Studies, 45–65. London: Longman. Barker, Hannah. 2002. England, 1760–1815. In: Hannah Barker and Simon Burrows (eds.), Press, Politics and the Public Sphere in Europe and North America, 1760–1820, 93–112. Cambridge: Cambridge University Press. Bazerman, Charles. 1988. Shaping Written Knowledge: The Genre and Activity of the Experimental Article in Science. Madison, WI: University of Wisconsin Press. Beal, Joan C. 2004. English in Modern Times 1700–1945. London: Arnold. Bell, Allan. 2003. Poles apart: Globalization and the development of news discourse across the twentieth century. In: Aitchison and Lewis (eds.), 7–17.
56 Late Modern English: Pragmatics and discourse Biber, Douglas. 1988. Variation across Speech and Writing. Cambridge: Cambridge University Press. Biber, Douglas. 2003. Compressed noun-phrase structures in newspaper discourse: The competing demands of popularization vs. economy. In: Aitchison and Lewis (eds.), 169–181. (Reprinted in Wolfgang Teubert and Ramesh Krishnamurthy (eds.), Corpus Linguistics: Critical Concepts in Linguistics. Vol. V, 130–141. London: Routledge.) Biber, Douglas. 2004. Historical patterns for the grammatical marking of stance. Journal of Historical Pragmatics 5(1): 107–136. Biber, Douglas and Edward Finegan. 1989. Drift and the evolution of English style: A history of three genres. Language 65(3): 487–517. Bijkerk, Annemieke. 2004. Yours sincerely and yours affectionately: On the origin and development of two positive politeness markers. Journal of Historical Pragmatics 5(2): 297–311. Brinton, Laurel J. 2001a. Historical discourse analysis. In: Deborah Schiffrin, Deborah Tannen, and Heidi E. Hamilton (eds.), The Handbook of Discourse Analysis, 138–160. Oxford: Blackwell. Brinton, Laurel J. 2001b. From matrix clause to pragmatic marker: The history of look-forms. Journal of Historical Pragmatics 2(2): 177–199. Brinton, Laurel J. 2007. What’s more: The development of pragmatic markers in the modern period. In: Pe´rez Guerra et al. (eds.), 47–75. Bern: Peter Lang. Brinton, Laurel J. 2008. The Comment Clause in English: Syntactic Origins and Pragmatic Development. Cambridge: Cambridge University Press. Dossena, Marina and Charles Jones (eds.). 2003. Insights into Late Modern English. Bern: Peter Lang. Dossena, Marina and Ingrid Tieken-Boon van Ostade (eds.). 2008. Studies in Late Modern English Correspondence. Methodology and Data. Berlin: Peter Lang. Fairman, Tony. 2006. Words in English Record Office documents of the early 1800s. In: Kyto¨ et al. (eds.), 56–88. Finell, Anne. 1992. The repertoire of topic changers in personal, intimate letters: A diachronic study of Osborne and Woolf. In: Matti Rissanen, Ossi Ihalainen, Terttu Nevalainen, and Irma Taavitsainen (eds.), History of Englishes: New Methods and Interpretations in Historical Linguistics, 720–735. Berlin/New York: Mouton de Gruyter. Finegan, Edward. 1992. Style and standardization in England 1700–1900. In: Timothy W. Machan and Charles T. Scott (eds.), English in its Social Context: Essays in Historical Sociolinguistics, 102–130. Oxford: Oxford University Press. Finegan, Edward. 1998. English grammar and usage. In: Romaine (ed.), 536–588. Fitzmaurice, Susan. 2007. The world of the periodical essay: Social networks and discourse communities in eighteenth-century London. Historical Sociolinguistics and Sociohistorical Linguistics 7, http://www.let.leidenuniv.nl/hsl_shl/periodicalessay.htm Fitzmaurice, Susan and Irma Taavitsainen (eds.). 2007. Methods in Historical Pragmatics. Berlin/ New York: Mouton de Gruyter. Fowler, Henry W. and Frank G. Fowler. 1924 [1919]. The King’s English. 2nd edn. Oxford: The Clarendon Press. Geisler, Christer. 2002. Investigating register variation in nineteenth-century English: A multidimensional comparison. In: Randy Reppen, Susan Fitzmaurice, and Douglas Biber (eds.), Using Corpora to Explore Variation, 249–271. Amsterdam/Philadelphia: John Benjamins. Gieszinger, Sabine. 2000. Two hundred years of advertising in The Times. In: Friedrich Ungerer (ed.), English Media Texts Past and Present: Language and Textual Structure, 85–109. Amsterdam/Philadelphia: John Benjamins. Go¨rlach, Manfred. 2001. Eighteenth-century English. Heidelberg: Winter. Go¨rlach, Manfred. 2002. A linguistic history of advertising 1700–1890. In: Teresa Fanego, Bele´n Me´ndez-Naya, and Elena Seoane (eds.) Sounds, Words, Texts and Change: Selected Papers from 11 ICEHL, Santiago de Compostela 7–11 September 2000, 83–104. Amsterdam/ Philadelphia: John Benjamins.
913
914
VI Late Modern English Hoffmann, Sebastian. 2006. Tag questions in Early and Late Modern English: Historical description and theoretical implications. Anglistik 17(2): 35–55. Hudson, Nicholas. 1994. Writing and European Thought 1600–1830. Cambridge: Cambridge University Press. Jucker, Andreas H. 2005. News discourse: Mass media communication from the seventeenth to the twenty-first century. In: Janne Skaffari, Matti Peikola, Ruth Carrol, Risto Hiltunen, and Brita Wa˚rvik (eds.), Opening Windows on Texts and Discourses of the Past, 7–21. Amsterdam/ Philadelphia: John Benjamins. Jucker, Andreas H. 2006. Historical pragmatics. In: Keith Brown (ed.), Encyclopedia of Language and Linguistics. 2nd edn., 329–331. Oxford: Elsevier. Jucker, Andreas H., Gerd Fritz, and Franz Lebsanft (eds.). 1999. Historical Dialogue Analysis. Amsterdam/Philadelphia: John Benjamins. Jucker, Andreas H. and Irma Taavitsainen (eds.). 2008. Speech Acts in the History of English. Amsterdam/Philadelphia: John Benjamins. Knowles, Gerry. 1997. A Cultural History of the English Language. London: Edward Arnold. Kyto¨, Merja, Juhani Rudanko, and Erik Smitterberg. 2000. Building a bridge between the present and the past: A corpus of nineteenth-century English. ICAME Journal 24: 85–97. Kyto¨, Merja, Mats Ryde´n, and Erik Smitterberg (eds.). 2006. Nineteenth-century English: Stability and Change. Cambridge: Cambridge University Press. Lewis, Diana M. 2003. Rhetorical motivations for the emergence of discourse particles, with special reference to English of course. Belgian Journal of Linguistics 16: 79–91. McIntosh, Carey. 1998. The Evolution of English Prose 1700–1800: Style, Politeness and Print Culture. Cambridge: Cambridge University Press. Nevala, Minna. 2007. Inside and out: Address forms in 17th- and 18th-century letters. In: Terttu Nevalainen and Sanna-Kaisa Tanskanen (eds.), Letter Writing, 89–113. Amsterdam/ Philadelphia: John Benjamins. Nevala, Minna. 2009. Altering distance and defining authority: Person reference in Late Modern English. Journal of Historical Pragmatics (Special issue on Historical Sociopragmatics, ed. by Jonathan Culpeper.) 10(2): 238–258 Oldireva Gustafsson, Larisa. 2006. The passive in nineteenth-century scientific writing. In: Kyto¨ et al. (eds.), 110–135. Orduna Nocito, Elena. 2007. The semantic field of “manners” in the eighteenth century: A cognitive approach. In: Pe´rez Guerra et al. (eds.), 383–400. Palander-Collin, Minna and Minna Nevala. 2010. Reporting and social role construction in eighteenth-century personal correspondence. In Pa¨ivi Pahta, Minna Nevala, Arja Nurmi, and Minna Palander-Collin (eds.), Social Roles and Language Practices in Late Modern English, 111–133. Amsterdam/Philadelpia: John Benjamins. Pe´rez Guerra, Javier, Dolores Gonza´lez Alvarez, Jorge L. Bueno Alonso, and Esperanza Rama Martı´nez (eds.). 2007. “Of Varying Language and Opposing Creed”: New Insights into Late Modern English. Bern: Peter Lang. Romaine, Suzanne (ed.). 1998. Cambridge History of the English Language. Vol. IV. 1776–1997. Cambridge: Cambridge University Press. Salager-Meyer, Francoise and Nahirana Zambrano. 2001. The bittersweet rhetoric of controversiality in nineteenth- and twentieth-century French and English medical literature. Journal of Historical Pragmatics 2(1): 141–174. Smitterberg, Erik. 2005. The Progressive in Nineteenth-century English. Amsterdam: Rodopi. Taavitsainen, Irma and Andreas Jucker. 2007. Historical corpus pragmatics: Methodology and case studies. Paper presented at the Tenth International Pragmatics Association Conference, Go¨teborg 8–13 July, 2007. Tieken-Boon van Ostade, Ingrid and Fa´tima Marı´a Faya Cerqueiro. 2007. Saying please in Late Modern English. In: Pe´rez Guerra et al. (eds.), 421–444.
57 Late Modern English: Dialects
915
Traugott, Elizabeth Closs. 2003. Historical pragmatics. In: Laurence R. Horn and Gregory L. Ward (eds.), The Handbook of Pragmatics, 538–561. Oxford: Blackwell. Valle, Ellen. 2004. “A nice and accurate philosopher”: Interactivity and evaluation in a historical context. In: Gabriella del Lungo Camiciotti and Elena Tognini Bonelli (eds.), Academic Discourse: New Insights into Evaluation, 55–80. Bern: Peter Lang. Watts, Richard. 2002. From polite language to educated language: The re-emergence of an ideology. In: Richard Watts and Peter Trudgill (eds.), Alternative Histories of English, 155–172. London: Routledge.
Diana M. Lewis, Aix-en-Provence (France)
57 Late Modern English: Dialects 1 2 3 4 5 6 7 8
Preliminaries A general assessment – English (dialects) in the 18th and 19th century Limits to studying non-standard morphosyntax Alexander Ellis’s On Early English Pronunciation (1889) Existing catalogues of Late Modern English dialect features Past and present contexts of selected morphosyntactic dialect features Summary References
Abstract Studying Late Modern English dialects is difficult for a number of reasons. This chapter will briefly discuss the status of English (dialects) in the Late Modern period before focussing on the limits to studying non-standard morphosyntax in particular. While contemporary descriptions tended to focus on the presence or absence of features, modern studies emphasize degrees and tendencies rather than either/ors. On Early English Pronunciation (Ellis 1869–89) may be used to show how traditional materials can be modified in order to establish limits of regional variation. The areas discussed are necessarily restricted to high-frequency phenomena such as pronominal usage and relative markers.
1 Preliminaries The Late Modern English (LModE) period is a fairly recent addition to the landscape of historical linguistics. While it may not be as “interesting” as the preceding periods with their chaos of emerging and/or completely re-organized systems, it holds its own against Old, Middle, or Early Modern English for a different reason: Late Modern English is probably the earliest period for which close-to-naturalistic speech data is available. Despite the fact that sound recordings are not on hand for obvious reasons, recent and future additions to the corpus-linguistic landscape such as the digitization of the Old Bailey Corpus (Huber and Maiwald 2003–10; see also Huber 2007) or the University of Innsbruck’s Speed project, which involves the digitization of Wright’s English Alex Bergs and Laurel Brinton 2012, Historical Linguistics of English (HSK 34.1), de Gruyter, 915–938
916
VI Late Modern English Dialect Dictionary (Markus and Heuberger 2011–) take an intermediate stage between spoken and written, vernacular, and standard-language corpora. In the long run, modern technology will enable detailed searches at an unprecedented scale in these large corpora. They will open the window to the past a bit further than (slightly) younger resources such as the Survey of English Dialects (Orton and Dieth [eds.] 1962–71), offering means to researchers to investigate features of traditional dialects with state-of-the-art research methods (including, for the first time ever, quantitative statements that not only take occurrences but also non-occurrences, i.e. the standard forms, into account). However, this bright-looking future should not lead to ignorance of the problems one is faced with when wanting to study Late Modern English dialects right now. Many of these problems will not disappear once these new resources become available – they involve issues of representativeness, authenticity, and careful assessment of the “facts”. In the following, some of these difficulties will be discussed in more detail, focusing on a number of (high-frequency) morphosyntactic features to exemplify methods to come to terms – if not solve – some of the problems. Please note that there is no agreement among scholars when exactly the Early Modern English period ends and the Late Modern one starts. Possible dates include 1776 (American secession), 1800 (round figure), 1815 (end of Napoleon’s reign), and 1832 (cheap printing becomes available; cf. Go¨rlach 1999b: 463).
2 A general assessment – English (dialects) in the 18th and 19th century As already indicated in the introduction, the 18th and 19th centuries are very different from the preceding centuries in the history of English from a linguistic point of view. The radical changes the (grammatical) system had been undergoing in the Middle and (though much less so) the Early Modern English period have largely been completed, as witnessed by the following statements: The eighteenth century inherited a largely ordered grammar from Early Modern English […] (Go¨rlach 1999b: 484). Since relatively few categorical losses or innovations have occurred [from the late 18th century onwards], syntactic change has more often been statistical in nature, with a given construction occurring throughout the period and either becoming more or less common generally or in particular registers. The overall, rather elusive effect can seem more a matter of stylistic than syntactic change (Denison 1998: 93).
As a result, attention in contemporary publications on the (state of the) English language shifted to new areas, and one area in particular: standardization in the form of (prescriptive) codification (cf. e.g. Beal 2004b: 329; Rissanen 1999: 211; Go¨rlach 1999b: 462, 482; see also Auer, Chapter 58; Lange, Chapter 62). For the first time in the history of the language, both pronunciation dictionaries and grammars were published that were intended for average people. While in earlier centuries Latin grammar had been the ideal towards which English grammar was supposed to orient itself, the 18th and 19th century gave way to a more realistic position. Go¨rlach (1999b: 483) lists a number of patterns that were established in that period which were not modeled on (and in fact had no parallel in) Latin grammar, including for example the
57 Late Modern English: Dialects development of the purely syntactic uses of do, the fixed word order, and the consolidation of aspectual distinctions. Given the mostly prescriptive attitude of the time, regional accents and dialects, if mentioned at all, are generally only discussed in connection with avoidance strategies. Beal (2004a, 2004b) gives an overview of the general tenor in such publications as Walker’s and Sheridan’s pronouncing dictionaries. While they provide a wealth of information for researchers interested in historical phonology, dialectologists will find little of interest in them. To the former group, pronouncing dictionaries of the time are particularly useful since they were intended for the London middle class, who were told to avoid a Cockney (-like) accent at all costs (cf. Beal 2004b: 331). With Cockney thus set as the “don’t” norm, conclusions can be drawn about regional pronunciations at the time. Beal (2004a: 192–209) shows how a close reading of such sources can indeed be used to trace such allegedly “modern” changes as TH-fronting (i.e. the pronunciation of